Energy Management

Power Electronics and
Energy Management
for Battery Storage
Systems
Edited by
Andrei Blinov and Sheldon Williamson
Printed Edition of the Special Issue Published in Energies
www.mdpi.com/journal/energies
Power Electronics and Energy
Management for Battery Storage
Systems
Power Electronics and Energy
Management for Battery Storage
Systems
Editors
Andrei Blinov
Sheldon Williamson
MDPI • Basel • Beijing • Wuhan • Barcelona • Belgrade • Manchester • Tokyo • Cluj • Tianjin
Editors
Andrei Blinov Sheldon Williamson
Tallinn University of Technology University of Ontario Institute of Technology
Estonia Canada
Editorial Office
MDPI
St. Alban-Anlage 66
4052 Basel, Switzerland
This is a reprint of articles from the Special Issue published online in the open access journal Energies
(ISSN 1996-1073) (available at: https://www.mdpi.com/journal/energies/special issues/peem for
battery storage systems).
For citation purposes, cite each article independently as indicated on the article page online and as
indicated below:
LastName, A.A.; LastName, B.B.; LastName, C.C. Article Title. Journal Name Year, Volume Number,
Page Range.
ISBN 978-3-0365-5277-4 (Hbk)

ISBN 978-3-0365-5278-1 (PDF)
© 2022 by the authors. Articles in this book are Open Access and distributed under the Creative
Commons Attribution (CC BY) license, which allows users to download, copy and build upon
published articles, as long as the author and publisher are properly credited, which ensures maximum
dissemination and a wider impact of our publications.
The book as a whole is distributed by MDPI under the terms and conditions of the Creative Commons
license CC BY-NC-ND.
Contents
About the Editors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
Omar Abdel-Rahim, Andrii Chub, Andrei Blinov, Dmitri Vinnikov and

Dimosthenis Peftitsis
An Efficient Non-Inverting Buck-Boost Converter with Improved Step Up/Down Ability
Reprinted from: Energies 2022, 15, 4550, doi:10.3390/en15134550 . . . . . . . . . . . . . . . . . . . 1
Sumukh Surya, Akash Samanta, Vinicius Marcis and Sheldon Williamson

Smart Core and Surface Temperature Estimation Techniques for Health-Conscious Lithium-Ion
Battery Management Systems: A Model-to-Model Comparison
Reprinted from: Energies 2022, 15, 623, doi:10.3390/en15020623 . . . . . . . . . . . . . . . . . . . . 19
Zbigniew Rymarski, Krzysztof Bernacki and Łukasz Dyga

Controlled Energy Flow in Z-Source Inverters
Akash Samanta and Sheldon S. Williamson

A Comprehensive Review of Lithium-Ion Cell Temperature Estimation Techniques Applicable
to Health-Conscious Fast Charging and Smart Battery Management Systems
Rafał Kopacz, Michał Harasimczuk, Bartosz Lasek, Rafał Miśkiewicz and Jacek Rąbkowski
All-SiC ANPC Submodule for an Advanced 1.5 kV EV Charging System under Various
Modulation Methods
Ilya A. Galkin, Andrei Blinov, Maxim Vorobyov, Alexander Bubovich, Rodions Saltanovs
and Dimosthenis Peftitsis
Interface Converters for Residential Battery Energy Storage Systems: Practices, Difficulties and
Prospects
Van Quan Dao, Minh-Chau Dinh, Chang Soon Kim, Minwon Park, Chil-Hoon Doh,
Jeong Hyo Bae, Myung-Kwan Lee, Jianyong Liu and Zhiguo Bai
Design of an Effective State of Charge Estimation Method for a Lithium-Ion Battery Pack Using
Extended Kalman Filter and Artificial Neural Network
Andres Annuk, Wahiba Yaı̈ci, Matti Lehtonen, Risto Ilves, Toivo Kabanen and Peep Miidla
Simulation of Energy Exchange between Single Prosumer Residential Building and Utility Grid
v
About the Editors
Andrei Blinov
Andrei Blinov received an M.Sc. degree in electrical drives and power electronics and a Ph.D.
degree, with a dissertation devoted to the research into switching properties and performance
improvement methods for high-voltage IGBT-based dc–dc converters, from Tallinn University of
Technology, Tallinn, Estonia, in 2008, and 2012, respectively. After his Ph.D. studies, he spent two
years in Sweden working as a Postdoctoral Researcher with the KTH Royal Institute of Technology.
He is currently a Senior Researcher with the Department of Electrical Power Engineering and
Mechatronics, Tallinn University of Technology. His research interests are research into switch-mode
power converters, new semiconductor technologies, and energy storage systems.
Sheldon Williamson
Sheldon Williamson received his Bachelors of Engineering (B.E.) degree in Electrical Engineering
with high distinction from University of Mumbai, Mumbai, India, in 1999. He received a Master’s of
Science (M.S.) degree in 2002, and the Doctor of Philosophy (Ph.D.) degree (with Honors) in 2006,
both in Electrical Engineering, from the Illinois Institute of Technology, Chicago, IL, specializing
in automotive power electronics and motor drives, at the Grainger Power Electronics and Motor
Drives Laboratory. Currently, Dr. Williamson is a Professor at the Smart Transportation Electrification
and Energy Research (STEER) group, within the Department of Electrical, Computer, and Software
Engineering, at Ontario Tech University, in Oshawa, Ontario, Canada. He also holds the prestigious
NSERC Canada Research Chair position in Electric Energy Storage Systems for Transportation
Electrification. His main research interests include advanced power electronics and motor drives
for transportation electrification, electric energy storage systems, and electric propulsion. Prof.
Williamson is a Fellow of the IEEE.
vii
energies
Article
An Efficient Non-Inverting Buck-Boost Converter with
Improved Step Up/Down Ability
Omar Abdel-Rahim 1,2, *, Andrii Chub 1 , Andrei Blinov 1 , Dmitri Vinnikov 1 and Dimosthenis Peftitsis 3
1 Department of Electrical Power Engineering and Mechatronics, Tallinn University of Technology,

19086 Tallinn, Estonia; andrii.chub@taltech.ee (A.C.); andrei.blinov@taltech.ee (A.B.);
dmitri.vinnikov@taltech.ee (D.V.)
2 Electrical Engineering Department, Aswan University, Aswan 81542, Egypt
3 Department of Electrical Power Engineering, Norwegian University of Science and Technology,
NO-7491 Trondheim, Norway; dimosthenis.peftitsis@ntnu.no
* Correspondence: o.abdelrahim@aswu.edu.eg
Abstract: In this article, a new non-inverting buck-boost converter with superior characteristics in
both bucking and boosting is presented. The proposed converter has some distinct features, such
as high step-up/-down ability and low voltage/current stress on its switching devices. The voltage
gain of the proposed converter is double the reported value for the traditional buck-boost converter.
Although it has three switches, the three switches operate simultaneously, hence no dead-time is
required. Two out of the three switches are under voltage stress equal to half of the output voltage.
The overall efficiency of the system is promising because of the ability to select devices with low
voltage drops. Converter analysis and steady-state performance in both continuous conduction
mode (CCM) and discontinuous conduction mode (DCM) are presented in detail. A 1 kW hardware
prototype of the converter was implemented in the laboratory; with a step-up ratio of 3.5 and 1 kW
power, the measured efficiency is above 95.4%, and with step-up ratio 8, it is around 91.5%.
Citation: Abdel-Rahim, O.; Chub, A.;
Blinov, A.; Vinnikov, D.; Peftitsis, D. Keywords: high-gain non-inverting buck-boost converter; continuous conduction mode (CCM);
An Efficient Non-Inverting Buck- discontinuous conduction mode (DCM)
Boost Converter with Improved Step
Up/Down Ability. Energies 2022, 15,
4550. https://doi.org/10.3390/
en15134550 1. Introduction
Academic Editors: Alon Kuperman Traditional buck-boost converters, CUK and SEPIC, are able to buck or boost input
and Marco Pau voltage; however, their bucking or boosting abilities are limited, and they have high stress
on their switching devices, hence their efficiency and applications are limited [1–6]. In order
Received: 17 May 2022
to improve step-up/-down abilities, a group of power converters have been developed
Accepted: 20 June 2022
in the literature [7–17]. The topology proposed in [7] is a modification of the traditional
Published: 22 June 2022
buck-boost converter with improved voltage gain, but it has an inverted output and two of
Publisher’s Note: MDPI stays neutral the switching devices are under high voltage stress. A high gain with continuous input
with regard to jurisdictional claims in current buck-boost converter has been proposed in [7,8], but the converter is inverted and
published maps and institutional affil- includes many storage devices. In [9], a novel buck-boost converter is proposed with lower
iations. component stresses and less storage devices, However, the converter has limited voltage
gain; high ripple; and the converter switches operate in a complementary manner, which
increases dead-time and switching protection issues.
The quadratic voltage gain buck-boost converters developed in [10–12] provide good
Copyright: © 2022 by the authors.
performance in step-up mode, but their step-down ability is very limited.
Licensee MDPI, Basel, Switzerland.
In [13,14], semi-quadratic buck-boost converters are proposed. Despite their improved
This article is an open access article
performance in both bucking and boosting modes, there is no common ground and the
distributed under the terms and
conditions of the Creative Commons
input current is discontinuous. A quasi-Y source-based buck-boost dc–dc converter is
Attribution (CC BY) license (https://
introduced in [15,16]. This converter achieved a very high voltage gain using two inductors.
creativecommons.org/licenses/by/
Nevertheless, the severe slope of the voltage gain ratio makes controlling the converter
4.0/). very difficult.
Energies 2022, 15, 4550. https://doi.org/10.3390/en15134550 1 https://www.mdpi.com/journal/energies

Energies 2022, 15, 4550
In order to achieve higher voltage and gain and sustain higher efficiency at a wide
range of input voltage change, this paper presents a new high-gain non-inverting buck-
boost converter. The structure proposed has different merits, such as non-inverting, high
voltage gain, reduced components’ stresses, and the ability to sustain better efficiency at
wide voltage and load ranges.
The rest of the paper is organized as follows: Section 2 discusses the principle of
operation and analysis of the proposed converter; Section 3 presents the experimental
results of the converter; and finally, Section 4 presents the conclusions.
2. Proposed Buck-Boost Dc–Dc Converter

The configuration of the proposed buck-boost dc–dc converter is illustrated in
Figure 1 [17]. The structure is implemented using three power switches (S1 , S2 , S3 ), two diodes
(D1 , Do ), two inductors (L1 , L2 ), and an output capacitor (Co ). The three switches are trig-
gered on and off simultaneously, and the diodes operate as freewheeling diodes. The two
inductors charge in parallel and discharge in series.
DR
S L S
D Co Ro Vo
VDC
L
S
Figure 1. Proposed buck-boost dc–dc converter.
The converter is able to operate in continuous conduction mode (CCM) and discontin-
uous conduction mode (DCM). Both modes of operation will be considered in the following
sections.
2.1. Continuous Conduction Mode

In order to simplify the analysis of the CCM mode, two assumptions are considered in
the forthcoming analysis:
Capacitor voltage ripple is very small compared with the voltage itself, thus it could
be neglected.
Inductor current ripple is negligible because of its very small value.
All semiconductor devices are ideal.
The converter power switches are triggered ON and OFF simultaneously, hence the
converter will have two operating modes; see Figure 2a,b. Typical waveforms of the
converter in CCM are shown in Figure 3.
Mode I [0-DTS ]: In this time period, switches (S1 , S2 , S3 ) are turned ON, while diodes
(D1 , Do ) are turned OFF. This mode is illustrated in Figure 2a. As can be seen from the
figure, the two inductors charge in parallel from the source. Applying Kirchhoff voltage law
(KVL) and Kirchhoff current law (KCL) to Figure 2a, the following equations are deduced:
v L1 = v L2 = Vdc (1)
Vo
ic = − (2)
R
id = 0 (3)
2
Energies 2022, 15, 4550
DR
S L S
D Co Ro Vo
VDC
S L
(a)
DR
S L S
VDC Co Ro Vo
D
S L
(b)
DR
S L S
D
Co Ro Vo
VDC
S L
(c)
Figure 2. Operation modes of the proposed converter: (a) operation mode #1, (b) operation mode #2,
and (c) operation mode #3.
Mode II [DTS -TS ]: In this time period, switches (S1 , S2 , S3 ) are turned OFF and,
consequently, diodes (D1 , Do ) are turned ON to provide a freewheeling path for the current.
This mode is illustrated in Figure 2b. As can be investigated from the figure, the two
inductors discharge their energies to the load in series. Applying Kirchhoff voltage law
(KVL) and Kirchhoff current law (KCL) to Figure 2b, the following equations are deduced:
v L = v L1 = v L2 (4)
2v L = −Vo (5)
Vo
iC = IL − (6)
R
id = IL (7)
The steady-state voltage gain of the proposed converter could be deduced from the
analysis of the two modes of operation by applying voltage second balance, and the voltage
gain of the proposed converter is as follows:
Vo 2D
=M= (8)
Vdc (1 − D )
3
Energies 2022, 15, 4550
where v L1 , v L2 , Vdc , Vo , IL , iC , id , M, and D are inductor L1 voltage, inductor L2 voltage,

input voltage, output voltage, inductor current, capacitor current, diode current, voltage
gain, and duty cycle, respectively.
t
Vͳ
Iͳ I6BSN āI/ V6BSN VGF
t
Vʹ
Iʹ I6BSN I/ V6BSN āVR
t
Vʹ
I͵ I6BSN I/ V6BSN āVR
t
Vͳ
Iͳ I'BSN I/ V'BSN VGF
t
V
I I'RBSN I/ V'RBSN VRVGF
t
I/BSN VGF/LāDT6 I/BSN VGF/LāDT6
Iͳ
Iʹ
t
DTs Ts
Figure 3. Typical converter waveforms in CCM.
2.2. Discontinuous Conduction Mode

The discontinuous conduction mode typically occurs with large inductor current
ripple in a converter operating at light load and containing current unidirectional switches.
However, some converters are purposely designed to operate in DCM. The proposed
converter will have three modes of operation while operating in DCM; see Figure 2a–c. The
typical converter waveform while operating in DCM is illustrated in Figure 4.
Mode I and Mode II, which were discussed in the previous section, are similar to
CCM analysis.
Mode III: In this interval, both the power switches and diodes are turned off. The
inductors’ currents are zero, as illustrated in Figure 2c.
vL = 0 (9)
Vo
iC = (10)
R
id = 0 (11)
Applying inductor volt-second balance in Equations (1), (5), and (9), the relation
between input and output voltage is obtained:
D1 ∗ vdc = D2 ∗ vo (12)
4
Energies 2022, 15, 4550
t
I6BSN āI/ V6BSN VGF
Vͳ
Iͳ
t
Vʹ I6BSN I/ V6BSN āVR
Iʹ
t
Vʹ I6BSN I/ V6BSN āVR
I͵
t
Vͳ
Iͳ I'BSN I/ V'BSN VGF
t
V
I I'RBSN I/ V'RBSN VRVGF
t
Iͳ I/BSN VGF/LāDT6 I/BSN VGF/LāDT6
Iʹ
t
Ts
D1Ts D2Ts
Figure 4. Typical converter waveforms in DCM.
The duty cycle D2 is an unknown, so a second equation is needed to eliminate D2 .

Capacitor charge-balance is used to obtain the second equation. The average of the diode
current is equal to the output current:
Vo
id = (13)
R
A sketch of the inductor and diode currents in DCM is illustrated in Figure 5a,b. The
dc component of the diode current is given by
T
1 S
id = id (t)dt (14)
TS 0
ISN VGF/L· DT6 ISN VGF/L· DT6

,/ W ,G W
t t
DT6 DT6 T6 DT6 DT6
D E
Figure 5. DCM operation (a) inductor current and (b) diode current.
5
Energies 2022, 15, 4550
The peak diode current could be obtained from the graph as
Vo
i pk = ∗ D1 ∗ TS (15)
2L
Solving Equations (13)–(15), the second required equation is obtained as
1 Vo Vo
id = ∗ D2 ∗ TS ∗ ∗ D1 = (16)
2 2L R
Let
2L
k= (17)
RTS
Then
D2 = 2 ∗ k/D1 (18)
Finally, the converter voltage gain in DCM operation is given as
Vo D12
=M= (19)
Vdc (2 ∗ K )
where D1 , D2 , TS , and R0 are periods when the switches are conducting, periods when the
diode is conducting, switching time, and load resistance, respectively.
The boundary for CCM and DC operation can be obtained by relating inductor current
and inductor ripple
IL > Δi L For CCM (20)
IL < Δi L For DCM (21)
Substituting CCM solutions for IL and Δi L in (20)
2
Vdc 2D Vdc
∗ > ∗ DTS (22)
R 1−D 2L
Equation (22) could be rearranged to

2
2L 2D
= K > D∗ (23)
RTS 1−D
Hence 2
2D
Kcri = D ∗ (24)
1−D
where Kcri is the critical boundary between CCM and DCM.
According to the above analysis, the converter can operate on CCM or DCM based on
the operating conditions; in order to avoid such conditions, accurate design of the converter
must be considered. Figure 6 represents the boundary condition between CCM and DCM
at different duty cycles and different power while the output voltage is fixed at 350 V.
2.3. Switches’ and Diodes’ Voltage Stresses

Voltage and current stress are important parameters in designing and selecting circuit
parameters, and the proposed converter switching elements’ stress is discussed below.
Switches S1 , S2 , and S3 are triggered in a simultaneous manner, but their ratings are
different. The voltage stress of switch S1 is equal to
VdsS1 = Vdc (25)
6
Energies 2022, 15, 4550
Figure 6. CCM and DC boundary.
The current stress of switch S1 is given by
IS1 = 2 ∗ IL (26)
Switches S2 and S3 face similar voltage and current stress, as follows:
Vo
VdsS2 = VdsS3 = (27)
2
IS2 = IS3 = IL (28)
Diodes D1 and Do work as freewheeling diodes and are activated in complementary
manners to the switches. The voltage and current stress of both diodes are given by
Vd1 = Vdc (29)
Id1 = IL (30)
Vdo = Vdc + Vo (31)
IS2 = IL (32)
A depiction of the devices’ normalized voltage stresses with different voltage gain is
illustrated in Figure 7. In Figure 7, the voltage stress is normalized to the input voltage.
S S
1RUPDOL]HG9ROWDJH6WUHVV
Do

S D

9ROWDJH*DLQ
Figure 7. Circuit component voltage stress normalized to input voltage vs. converter voltage gain.
7
Energies 2022, 15, 4550
2.4. Components’ Design

The design of the circuit parameters, inductors, and capacitor is obtained from the
steady-state analysis performed in the previous sections. Utilizing inductor volt-second
balance and capacitor charge, the designs of parameters are as follows:
2.4.1. Inductors’ Design

Inductors’ selection is based on the required ripple of its current. The inductor current
ripple in CCM is drawn in Figure 8a and is given by
of f Vdc DTS Vo (1 − D ) TS
Δi L = Δion
L = Δi L = = (33)
L 2L
9F W
,/ W
&KDUJH 'LVFKDUJH &KDUJH

0 DT6 T6 DT6 T6 t
D E
Figure 8. CCM operation (a) inductor current and (b) output capacitor voltage.
This equation is valid in both CCM and DCM. By defining the required amount of
ripple, the inductor value could be defined as follows:
(Vo ∗ (1 − D )) ∗ TS
L= (34)
(2 ∗ Δi L )
Based on Equation (29), there is a dependency between inductance L and duty cycle D.
In order to avoid any misoperation of the converter, we design the inductance based on
the extreme condition that the current ripple at the extreme scenario does not exceed the
required ripple and when duty cycle below the ripple will be below the required level.
Let us assume the required ripple Δi L is 10%, then we can calculate L at the duty cycle
around 0.82. Then, when the duty cycle is lower than 0.85, the ripple will be less than 10%.
2.4.2. Capacitors’ Design

The output capacitor value is selected based on the amount of voltage ripple acceptable
in the output voltage. The output capacitor voltage waveform is illustrated in Figure 8b
and the ripple equation is as follows:
of f (Vo − IL)(1 − D ) TS Vo DTS

Δv = Δvon
c = Δvo = = (35)
RC RC
This equation is valid in both CCM and DCM. By defining the required amount of
ripple, the capacitor value could be defined as follows:
(Vo ∗ D ∗ TS )
Co = (36)
(Δv ∗ R)
of f of f
where Δi L , Δion
L , Δi L , L, Δv, Δvc , Δvo , and C are inductor current ripple, inductor
on
ripple while the inductor is charging, inductor ripple while the inductor is discharging,
inductor value, capacitor ripple, capacitor ripple while the capacitor is charging, capacitor
ripple while the capacitor is discharging, and capacitor value, respectively.
Based on Equation (30), there is a dependency between capacitance C and duty cycle
D. In order to avoid any misoperation of the converter, we design the capacitance based on
8
Energies 2022, 15, 4550
the extreme condition that the voltage ripple at the extreme scenario does not exceed the
required ripple and when duty cycle below the ripple will be below the required level
Let us assume the required ripple Δv is 10%, then we can calculate C at the duty cycle
around 0.82. Then, when the duty cycle is lower than 0.85, the ripple will be less than 10%.
2.4.3. Comparative Study

Converter performance mainly depends on the input voltage, input power, and step-
up ratio. The converter voltage gain is affected by the loading profile and supply voltage.
In Figure 9a, the input voltage is fixed at 150 V, while different loading profiles are applied;
at light load, the converter conversion ability is higher than at heavy loading. The second
case study is illustrated in Figure 9b, where load profile is fixed with different supply
voltages; as the supply voltage increases, the step-up/-down ability increases.
ͳͷͲǡͲǤͷ ε ͳͷͲǡͳ ε ͳͷͲǡͷ ε ͳͷͲǡ͵

ͳʹ ͳ ͸Ǥͷ
ͲǤͺ ͷǤͷ
ͳͲ ͲǤ͸
ͶǤͷ
ͲǤͶ

ͺ ͲǤʹ ͵Ǥͷ
ͲǤʹ ͲǤʹͶ ͲǤʹͺ ͲǤ͵ʹ ͲǤ͸͸ ͲǤ͹ ͲǤ͹Ͷ
͸
ͲǤʹ ͲǤ͵ ͲǤͶ ͲǤͷ ͲǤ͸ ͲǤ͹ ͲǤͺ

(a)
9: ε 9: ε 9: ε 9:

9ROWDJH*DLQ

'XW\&\FOH
'XW\&\FOH

'XW\&\FOH
(b)
9N: ε 9N: ε 9N: ε 9N:

(IILFLHQF\

9ROWDJH*DLQ
(c)
Figure 9. Different cases of study for the proposed converter: (a) Voltage gain at different duty cycle
and different loading. (b) Voltage gain with fixed loading and different input voltages. (c) Converter
efficiency at different input voltages.
9
Energies 2022, 15, 4550
However, in both case studies, the differences in the voltage gain do not have a very
high ratio.
Converter efficiency depends on many factors such as the load profile, source voltage,
and voltage gain. In the scenario illustrated in Figure 9c, the load profile is fixed while both
voltage gain and source voltage are variable. In buck mode, as the source voltage increases
and the bucking ratio is lowered, the converter demonstrates the highest efficiency, while
with lower input voltage and a higher bucking ratio, the converter efficiency is low. During
boosting mode, as source voltage increases, efficiency increases too.
Another case study is considered in Figure 10a, where input voltage is set to 150 V,
while load profile is variable and efficiency is measured at different voltage gains. With
heavy loading, the converter demonstrates lower efficiency than with a medium or moder-
ate loading profile. A comparison between the proposed converter and different converters
reported in the literature is illustrated in Table 1. In the voltage gain comparison illustrated
in Figure 10b, both the proposed converter and switched inductor buck-boost converter
have similar step-up/-down ability, but the proposed converter has higher efficiency; see
Figure 10c.
Table 1. Voltage gain and component stress comparison.
Converter Gain Components’ Count Switches’ and Diodes’ Switches’ and Diodes’
Topology M = Vo /Vin Switch Diode L C Voltage Stress Current Stress
S:Vo IL
Buck-Boost [18] (D/(1-D)) 1 1 1 1
D:(Vo +Vin ) IL
S1 :Vin IL
S2 :Vo IL
Non-Inverting [19] (D/(1-D)) 2 2 1 1
D1 :Vin IL
Do :Vo IL
S1 :Vin /(1-D) IL
Cuk [5] (D/(1-D)) 1 2 2 1 D1 :Vin IL
Do :Vo IL
S1 :Vo +Vin 2IL
SEPIC [19] (D/(1-D)) 1 1 2 2
D1 :Vo +Vin IL
S1 :Vo +Vin 2IL
D1 :Vin IL
SIBBC [20] 2D/(1-D) 1 4 2 1 D2 :Vo /2 IL
D3 :Vo /2 IL
Do :Vo +Vin IL
S1 :(Vo +Vin )/2 IL
S2 :(Vo +Vin )/2 IL
Lakshmi [21] (1+D1)/(1-D1 -D2 ) 3 2 2 1 S3 :Vo IL
D1 :Vo IL
Do :Vo +Vin IL
S1 :1/(1-D)*Vin IL
S2 :Vin *(1+D)/(1-D)2 IL
[22] 2D/(1-D)2 2 3 2 3 D1 :Vin /(1-D) IL
D2 :Vin /(1-D) IL
Do :Vo +Vin /(1-D) IL
S1 :1/(1-D)*Vin 2IL
S2 :Vin *D/(1-D)2 IL
[23] D2 /(1-D)2 2 2 2 2
D1 :Vin /(1-D) IL
D2 :Vin *D/(1-D)2 IL
10
Energies 2022, 15, 4550
Table 1. Cont.
Converter Gain Components’ Count Switches’ and Diodes’ Switches’ and Diodes’
Topology M = Vo /Vin Switch Diode L C Voltage Stress Current Stress
S1 :Vin 2IL
S2 :Vo /2 IL
Proposed 2D/(1-D) 3 2 2 1 S3 :Vo /2 IL
D1 :Vin IL
Do :Vo +Vin IL
9N: ε 9N: ε 9N: ε 9N:

%RRVW
%XFN

(IILFLHQF\

9ROWDJH*DLQ
9ROWDJH*DLQ

9ROWDJH*DLQ
(a)
3URSRVHG %XFN%RRVW 1RQ,QYHUWLQJ 6ZLWFKHG,QGXFWRU

9ROWDJH*DLQ

'XW\&\FOH

'XW\&\FOH
(b)
9LQ 93LQ :

%RRVW
%XFN

(IILFLHQF\

3URSRVHG&RQYHUWHU %XFNERRVW&RQYHUWHU
6ZLWFKHG,QGXFWRU%XFN%RRVW 1RQ,QYHUWLQJ%XFN%RRVW

9ROWDJHJDLQ
(c)
Figure 10. (a) Converter efficiency at fixed input voltage and different loading, (b) voltage gain
comparison among the proposed and other buck-boost converters, and (c) efficiency comparison
among the proposed and other buck-boost converters.
11
Energies 2022, 15, 4550
3. Experimental Verification
This section provides the experimental results of the developed system, and the
parameters used to build the prototype are illustrated in Table 2. A photo of the proposed
system is shown in Figure 11.
Table 2. Hardware prototype specifications.
Parameter Value
Input voltage range [V] 33–150
P [W] 700 W
Fs Switching Frequency 30 kHz
Switches S1 , S2 , S3 IMZ120R030M1HXKSA1
Diodes D1 , Do DPG10I300PA
Inductors L1 = L2 1 mH
Capacitor Co 320 μF

Figure 11. Experimental set-up schematic.
A case study where the duty cycle is set to 0.6 with 30 V input voltage is illustrated
in Figure 12. Three switches are operating in synchronous manner, hence the gate source
pulses for the three switches are the same as illustrated in Figure 12a. Diodes D1 and
Do are operating as freewheeling diodes. The cathode–anode voltages of the two diodes
are illustrated in Figure 12b. The input current is the sum of the two inductors’ currents
when the switches are on and zero when the switches are off, and the input capacitor
smoothens the input current. The drain source voltages of the three switches are illustrated
in Figure 12c. Switches S2 and S3 face the same voltage stress and carry the same current.
Figure 12d illustrates switch S1 current, which is equal to the sum of the two inductors’
currents. Switch S2 current is illustrated in Figure 12e, which is equal to the inductor
current. The output diode current is illustrated in Figure 12f, where spikes are noted in the
switches and diode currents because of a problem in the used probe; however, it does not
exist in real current as there are no spikes in the measured voltages.
A boosting case study is considered in Figure 13, where the input voltage is 25 V and
the output voltage generated is around 38 V, and a bucking case study is illustrated in
Figure 14, where the input voltage is 23 V, output voltage is 9.25 V, duty cycle is 0.2244, and
voltage gain is 0.4.
12
Energies 2022, 15, 4550
(a) (b)
(c) (d)
(e) (f)
Figure 12. Experimental results of converter at duty cycle of 0.6 and input voltage of 30 V: (a) gate
source pulses; (b) Ch1 diode D1 voltage, diode Do voltage, and input current; (c) Ch1 output diode
Do voltage, Ch3 switches’ S2 and S3 voltage, and Ch2 inductor L1 current; (d) Ch1 switch S3 voltage;
Ch2 switch S3 currents; (e) Ch2 switch S2 voltage, Ch3 switch S2 current, and inductor current; and
(f) Ch2 diode Do voltage and diode Do current.
Figure 13. Boosting case study, where the input voltage is 25 V, output voltage is 38 V, duty cycle is
0.428, and voltage gain is 1.52.
13
Energies 2022, 15, 4550
Figure 14. Bucking case study, where the input voltage is 23 V, output voltage is 9.25 V, duty cycle is
0.2244, and voltage gain is 0.4.
The converter voltage gain was measured experimentally, and the theoretical and
measured voltage gains of the converter with varying duty cycles are illustrated in Figure 15.
For comparison purposes, three prototypes were built in the laboratory for the traditional
buck-boost, non-inverting buck-boost, and proposed converter. The three prototypes were
built using the same parameters as in Table 2. In the first case study, which is illustrated
in Figure 16, the input voltage is set to 100 V and the step ratio is fixed at 3.7. For such a
step-up ratio, the proposed converter requires a duty cycle of 0.68, while the conventional
and non-inverting buck-boost converters both require a duty cycle of 0.8.
&DOFXODWHG 0HDVXUHG

9ROWDJH*DLQ

'XW\&\FOH
Figure 15. Calculated and measured converter voltage gain vs. duty cycle.
1RQ,QYHUWLQJ 7UDGLWLRQDO 3URSRVHG

9LQ 9 6WHSXSUDWLR

(IILFLHQF\

,QSXW3RZHU :
Figure 16. Measured efficiency comparison between the proposed converter, non-inverting buck-
boost, and traditional buck-boost converter at a step-up ratio of 3.7.
14
Energies 2022, 15, 4550
In this case study, the non-inverting converter demonstrates the highest efficiency,
while at high power, both the proposed and conventional converter have the same efficiency.
In the second case study, which is illustrated in Figure 17, the input voltage is fixed at 30 V
and the step-up ratio is 8. The efficiency of the proposed converter and the non-inverting
converter is comparable, but with the increase in power (over 300 W), the proposed
converter demonstrates the highest efficiency.
7UDGLWLRQDO 1RQ,QYHUWLQJ 3URSRVHG
9LQ 9 6WHSXSUDWLR

(IILFLHQF\

,QSXW3RZHU :
Figure 17. Measured efficiency comparison between the proposed, non-inverting, and traditional
buck-boost converter at a step-up ratio of 8.
The last case study demonstrates step-down comparison. In Figure 18, the input volt-
age is fixed at 150 V and the step-down ratio is 3. The non-inverting converter demonstrates
the lowest efficiency. The proposed converter and the conventional converter demonstrate
comparable efficiency at low power, but with the increase in input power, the proposed
converter demonstrates the highest efficiency.
&RQYHQWLRQDO 3URSRVHG 1RQ,QYHUWLQJ

9LQ 9 6WHSGRZQUDWLR

(IILFLHQF\

3RZHU :
Figure 18. Measured efficiency comparison between the proposed, non-inverting buck-boost, and
traditional buck-boost converter at a step-down ratio of 3.
4. Conclusions
In this paper, a new non-inverting high-gain buck-boost structure is developed with
improved step-up/-down ability. The performance of the converter in both CCM and
DCM is studied and analyzed. The design of the converter elements is investigated and
described. The operating conditions and voltage/current stress of each device are studied.
Based on the performed analysis, the proposed converter devices are under low voltage
and current stress compared with other buck-boost converters. A 700 W prototype was
built for the converter to investigate its performance experimentally. The efficiency of
the proposed converter is measured at different voltage gains and compared with the
traditional buck-boost converter.
15
Energies 2022, 15, 4550
The theoretical and measured voltage gain matched. While working in step-up, the
converter demonstrated better performance at high power. The peak measured efficiency
of the converter at a step-up ratio of 3.7 was 95.4%.
Author Contributions: O.A.-R. developed and simulated the idea; O.A.-R., A.C. and A.B. imple-
mented the idea prototype and verified the experimental results; O.A.-R. and D.V. revised the data
and results; O.A.-R. wrote the first draft; O.A.-R., D.V. and D.P. revised the manuscript. All authors
have read and agreed to the published version of the manuscript.
Funding: This research was financed in part by the European Economic Area (EEA) and Norway
Financial Mechanism 2014–2021 under Grant EMP474 and in part by the Estonian Research Council
under Grant PRG1086 and the programme Mobilitas Pluss (Grant MOBJD1033).
Conflicts of Interest: The authors declare no conflict of interest.
References
1. Abdel-Rahim, O.; Chub, A.; Blinov, A.; Vinnikov, D. Buck-Boost Resonant Z-Source Parital Power Converter. In Proceedings of
the 2022 3rd International Conference on Smart Grid and Renewable Energy (SGRE), Doha, Qatar, 20–22 March 2022; pp. 1–6.
[CrossRef]
2. Abdel-Rahim, O.; Orabi, M.; Ahmed, M.E. Buck-Boost Interleaved Inverter For Grid Connected Photovoltaic System. In
Proceedings of the 2010 IEEE International Conference on Power and Energy, Kuala Lumpur, Malaysia, 29 November–1 December
2010; pp. 63–68. [CrossRef]
3. Callegaro, L.; Ciobotaru, M.; Fletcher, J.E. An Intelligent Pass-Through Algorithm for Non-Inverting Buck-Boost Solar Power
Optimizers. In Proceedings of the 2019 21st European Conference on Power Electronics and Applications (EPE ’19 ECCE Europe),
Genova, Italy, 3–5 September 2019; pp. P.1–P.10. [CrossRef]
4. Linares, L.; Erickson, R.W.; MacAlpine, S.; Brandemuehl, M. Improved Energy Capture in Series String Photovoltaics via Smart
Distributed Power Electronics. In Proceedings of the 2009 Twenty-Fourth Annual IEEE Applied Power Electronics Conference
and Exposition, Washington, DC, USA, 15–19 February 2009; pp. 904–910. [CrossRef]
5. Abdel-Rahim, O.; Orabi, M.; Ahmed, M.E. High Gain Single-Stage Inverter For Photovoltaic AC Modules. In Proceedings of
the 2011 Twenty-Sixth Annual IEEE Applied Power Electronics Conference and Exposition (APEC), Fort Worth, TX, USA, 6–11
March 2011; pp. 1961–1967. [CrossRef]
6. Banaei, M.R.; Bonab, H.A.F. A High Efficiency Nonisolated Buck–Boost Converter Based on ZETA Converter. IEEE Trans. Ind.
Electron. 2019, 67, 1991–1998. [CrossRef]
7. Banaei, M.R.; Bonab, H.A.F. A novel structure for single-switch nonisolated transformerless buck–boost DC–DC con-verte. IEEE
Trans. Ind. Electron. 2016, 64, 198–205. [CrossRef]
8. Ajami, A.; Ardi, H.; Farakhor, A. Design, analysis and implementation of a buck–boost DC/DC converter. IET Power Electron.
2014, 7, 2902–2913. [CrossRef]
9. Son, H.-S.; Kim, J.-K.; Lee, J.-B.; Moon, S.-S.; Park, J.-H.; Lee, S.-H. A New Buck–Boost Converter With Low-Voltage Stress and
Reduced Conducting Components. IEEE Trans. Ind. Electron. 2017, 64, 7030–7038. [CrossRef]
10. Abdel-Rahim, O.; Furiato, H. Switched Inductor Quadratic Boosting Ratio Inverter With Proportional Resonant Controller For
Grid-Tie PV Applications. In Proceedings of the IECON 2014—40th Annual Conference of the IEEE Industrial Electronics Society,
Dallas, TX, USA, 29 October–1 November 2014; pp. 5606–5611. [CrossRef]
11. Maldonado, J.C.M.; Valdez-Resendiz, J.E.E.; Vite, P.M.G.; Rosas-Caro, J.C.; Rivera-Espinosa, M.D.R.; Valderrabano-Gonzalez, A.
Quadratic Buck–Boost Converter with Zero Output Voltage Ripple at a Selectable Operating Point. IEEE Trans. Ind. Appl. 2019,
55, 2813–2822. [CrossRef]
12. Rosas-Caro, J.C.; Valdez-Resendiz, J.E.; Mayo-Maldonado, J.C.; Alejo-Reyes, A.; Valderrabano-Gonzalez, A. Quadratic buck–boost
converter with positive output voltage and minimum ripple point design. IET Power Electron. 2018, 11, 1306–1313. [CrossRef]
13. Nousiainen, L.; Suntio, T. Dynamic characteristics of current-fed semi-quadratic buck-boost converter in photovoltaic applications.
In Proceedings of the 2011 IEEE Energy Conversion Congress and Exposition, Phoenix, AZ, USA, 17–22 September 2011; pp.
1031–1038. [CrossRef]
14. Yari, K.; Shahalami, S.H.; Mojallali, H. A Novel Nonisolated Buck–Boost Converter with Continuous Input Current and
Semiquadratic Voltage Gain. IEEE J. Emerg. Sel. Top. Power Electron. 2021, 9, 6124–6138. [CrossRef]
15. Yari, K.; Mojallali, H.; Shahalami, S.H. A New Coupled-Inductor-Based Buck–Boost DC–DC Converter for PV Applications. IEEE
Trans. Power Electron. 2021, 37, 687–699. [CrossRef]
16. AL-Hakeem, K.M.; Ahmad, O.A. Buck-Boost In-Out of Phase Y-source AC Matrix Converter. In Proceedings of the 2018 Third
Scientific Conference of Electrical Engineering (SCEE), Baghdad, Iraq, 19–20 December 2018; pp. 274–279. [CrossRef]
17. Abdel-Rahim, O.; Chub, A.; Blinov, A.; Vinnikov, D. New High-Gain Non-Inverting Buck-Boost Converter. In Proceedings of the
IECON 2021—47th Annual Conference of the IEEE Industrial Electronics Society, Toronto, ON, Canada, 13–16 October 2021; pp.
1–6. [CrossRef]
18. Smedley, K.; Cuk, S. Dynamics of one-cycle controlled Cuk converters. IEEE Trans. Power Electron. 1995, 10, 634–639. [CrossRef]
16
Energies 2022, 15, 4550
19. Ericson, R.; Maksimovic, D. Fundamentals of Power Electronics; Kluwer Academic Publishers: Dordrecht, The Netherlands, 2007.
20. Abdel-Rahim, O.; Orabi, M.; Abdelkarim, E.; Ahmed, M.; Youssef, M.Z. Switched Inductor Boost Converter For PV Applications.
In Proceedings of the 2012 Twenty-Seventh Annual IEEE Applied Power Electronics Conference and Exposition (APEC), Orlando,
FL, USA, 5–9 February 2012; pp. 2100–2106. [CrossRef]
21. Lakshmi, M.; Hemamalini, S. Nonisolated High Gain DC–DC Converter for DC Microgrids. IEEE Trans. Ind. Electron. 2017, 65,
1205–1212. [CrossRef]
22. Sarikhani, A.; Allahverdinejad, B.; Hamzeh, M. A Nonisolated Buck–Boost DC–DC Converter with Continuous Input Current for
Photovoltaic Applications. IEEE J. Emerg. Sel. Top. Power Electron. 2020, 9, 804–811. [CrossRef]
23. Miao, S.; Wang, F.; Ma, X. A New Transformerless Buck–Boost Converter with Positive Output Voltage. IEEE Trans. Ind. Electron.
2016, 63, 2965–2975. [CrossRef]
17
energies
Article
Smart Core and Surface Temperature Estimation Techniques for
Health-Conscious Lithium-Ion Battery Management Systems:
A Model-to-Model Comparison
Sumukh Surya 1, *, Akash Samanta 2 , Vinicius Marcis 2 and Sheldon Williamson 2
1 Robert Bosch Engineering and Business Solutions, Bangalore 560100, India

2 Department of Electrical, Computer and Software Engineering, Ontario Tech University,
Oshawa, ON L1G 0C5, Canada; akash.samanta@ontariotechu.net (A.S.);
vinicius.marcis@ontariotechu.net (V.M.); sheldon.williamson@uoit.ca (S.W.)
* Correspondence: sumukhsurya@gmail.com
Abstract: Estimation of core temperature is one of the crucial functionalities of the lithium-ion Battery
Management System (BMS) towards providing effective thermal management, fault detection and
operational safety. It is impractical to measure the core temperature of each cell using physical sensors,
while at the same time implementing a complex core temperature estimation strategy in onboard
low-cost BMS is also challenging due to high computational cost and the cost of implementation.
Typically, a temperature estimation scheme consists of a heat generation model and a heat transfer
model. Several researchers have already proposed ranges of thermal models with different levels of
accuracy and complexity. Broadly, there are first-order and second-order heat resistor–capacitor-based
thermal models of lithium-ion batteries (LIBs) for core and surface temperature estimation. This
paper deals with a detailed comparative study between these two models using extensive laboratory
test data and simulation study. The aim was to determine whether it is worth investing towards
developing a second-order thermal model instead of a first-order model with respect to prediction
Citation: Surya, S.; Samanta, A.;
accuracy considering the modeling complexity and experiments required. Both the thermal models
Marcis, V.; Williamson, S. Smart Core
and Surface Temperature Estimation
along with the parameter estimation scheme were modeled and simulated in a MATLAB/Simulink
Techniques for Health-Conscious environment. Models were validated using laboratory test data of a cylindrical 18,650 LIB cell.
Lithium-Ion Battery Management Further, a Kalman filter with appropriate process and measurement noise levels was used to estimate
Systems: A Model-to-Model the core temperature in terms of measured surface and ambient temperatures. Results from the
Comparison. Energies 2022, 15, 623. first-order model and second-order models were analyzed for comparison purposes.
https://doi.org/10.3390/en15020623
Keywords: electric vehicles; stationary battery energy storage system; battery automated system;
Academic Editor: Haifeng Dai
online state estimation; thermal modeling; first-order model; second-order model; Kalman filtering
Received: 14 November 2021
Accepted: 4 January 2022
Published: 17 January 2022
Publisher’s Note: MDPI stays neutral 1. Introduction

with regard to jurisdictional claims in Lithium-ion batteries (LIBs) have been extensively commercialized as a primary energy
published maps and institutional affil- storage technology for electric vehicles (EVs), stationary energy storage in the smart grid
iations. system and several other consumer electronics. The primary dominating factors of LIBs
over other energy storage technologies include high energy density, long lifespan, and de-
clining cost [1–4]. However, from literature and practice, it is noticed that the performance
of LIBs as well as the durability and reliability are significantly influenced by the operat-
ing temperature. Moreover, excessively high temperatures may cause thermal runaway,
leading to fire, smoke and other serious safety hazards to the operators [5–7]. Therefore,
the requirement of a battery management system (BMS) has become indispensable for
effective thermal management and safety of LIB system, which essentially requires accurate
information on the core and surface temperature of each cell [8,9] besides other important
creativecommons.org/licenses/by/ states such as state of charge (SOC) [10,11] and state of health (SOH). A few other popular
4.0/). functions of an advanced BMS include cell balancing [12,13], fault detection/diagnosis [14]

Energies 2022, 15, 623
and some other safety inspection functionalities. Several recent research studies highlighted
that the accuracy of estimating cell SOC [15], SOH [16] and remaining storage capacity [17]
depends on the accurate estimation of cell temperature as all these states are the function
of temperature. Moreover, the Columbic efficiency of a cell is greatly affected by the cell
temperature during the charging and discharging period. It is worthwhile to mention
that the temperature distribution inside the cell is not uniform, and the core temperature
remains higher than the surface temperature during practical application, especially under
high charging and discharging current [18]. Typically, the difference between the core and
surface temperature varies in the range of 5–10 ◦ C [19,20]; however, under high current
loading with rapid load fluctuation, it could be even more. Therefore, accurate informa-
tion on the core and surface temperature is essential to achieving the effective thermal
management of an LIB pack besides fault detection. While most of the existing temper-
ature measurement techniques measure the surface temperature directly using physical
sensors [21], the measurement of cells’ internal temperature is highly challenging when
using a physical sensor. Moreover, any high-capacity LIB pack consists of thousands of
single LIB cells; thus, installing physical sensors in each cell is not practically feasible from
the viewpoint of incremental cost and manufacturing complexity.
To sum up, accurate information on core temperature undoubtedly serves as the
essential basis for the thermal management and safety of LIB apart from SOC and SOH
estimation whilst it is difficult to measure the core temperature using physical sensors.
Therefore, a precise thermal model is crucial to accurate temperature estimation. Moreover,
it should be easy to model and computationally inexpensive in order to be implemented
in onboard BMS for online prediction of temperature. Several temperature estimation
techniques have been proposed by researchers. Typically, a temperature estimation strategy
consists of two models, namely, a heat generation model and a heat transfer model [22].
The heat generation model takes physical measurement signals from a cell, typically
voltage current, to estimate the total heat generation during charging and discharging.
Then, the heat transfer model takes the estimated total heat quantity as model input to
predict the temperature of that cell. Depending on the modeling, it can only estimate the
core temperature (single-state) or both the core and surface temperature simultaneously
(two-state).
Broadly, heat generation models can be classified into three groups, electrochemi-
cal models [23–26], data-driven empirical models [27–29] and equivalent circuit models
(ECM) [30–32]. Few other researchers have also grouped the heat generation model from
the perspective of heat concentration. According to them, the heat generation model could
be a concentrated model (all heat is generated at the core), distributed model (heat gener-
ated uniformly over the cell) [33] and heterogeneous model [30,34] (due to temperature and
current density gradient inside the cell). On the other hand, the heat transfer model can be
grouped into finite element analysis (FEA)-based models [32,35–38], lumped multi-node
models [27,39–41] and heat capacitive-resistive models [42]. The lumped multi-node model
and heat capacitive-resistive models are typically developed based on the analogy between
thermal and electrical phenomena. It can be seen that the electrochemical model can pro-
duce a very accurate heat generation value provided all model parameters are carefully
tuned. However, the electrochemical models are highly complex and computationally
expensive. The accuracy of data-driven empirical models highly depends on the experi-
mentally acquired data. Collecting such high-resolution data is challenging, and with the
increase in data volume and the number of feature vectors, computational expenses also in-
crease exponentially. On the other hand, an ECM-based estimation model can be designed
suitable for online prediction and real-world application by establishing a balance between
the computational cost and prediction accuracy. Therefore, ECM-based battery models are
extensively used in practice for estimating heat generation in LIB. Further, as far as the
heat transfer model is concerned, the heat resistor–capacitor models are easy to develop
and computationally efficient compared to FEA-based methods and lumped-parameter
multi-node models. The FEA-based methods are highly accurate; however, they come at
20
Energies 2022, 15, 623
the expense of a high computational cost. Resistor–capacitor-based models can be opti-

mally engineered to make a balance between prediction accuracy and computational cost
depending on the application requirement. Therefore, heat resistor–capacitor model-based
temperature estimation is the prime focus of this present study.
Researchers have proposed different kinds of heat resistor–capacitor models for the
accurate and precise internal and surface temperature estimation of LIB. However, the
major concern regarding the practical application of any model is its computational cost,
the capability of online prediction and suitability for onboard BMS. Detailed studies on
the thermal characteristics of different layers inside an LIB cell, modelling complexity
and the experimental data requirement have been carried out and are listed in the refer-
ences section [43–48]. The heat resistor–capacitor models use the analogy between thermal
and electrical phenomena, where heat capacity (thermal capacitance) and heat transfer
coefficient (thermal resistance) are represented as electrical capacitor and resistor, respec-
tively [43]. So far, a first-order (one thermal energy storage element) and second-order (two
thermal energy storage elements)-based thermal models have been reported in the litera-
ture for temperature estimation. Second-order models are typically complex and require
extensive experiments alongside the knowledge of domain experts during modeling. On
the other hand, first-order models are easy to implement, computationally inexpensive
and require far fewer experiments. Recently, extensive research effort has been made on
second-order thermal models of LIB. However, a comparative study between the first-order
and second-order model has not yet been assessed. Therefore, this research study focused
on the comparative study to investigate whether it is worth investing in developing and
implementing a second-order thermal model for the core temperature estimation of LIB in
terms of accuracy, modeling complexity and the experimental requirement and its practica-
bility in onboard BMS. Extensive experiments were conducted for data collection, and the
data was further utilized for modeling, validation and comparison purposes. The strategy
was to employ an ECM-based heat generation model for both a first-order and second-
order thermal model to determine the total heat generation inside the cell. A Kalman filter
(KF) was used in both the cases to improve prediction performance. Then, the estimation
results were compared with the measured data to assess the modeling accuracy. Finally,
the predicted results obtained from the first-order and second-order model were compared
for the purpose of model-to-model comparison.
The remaining portion of the article is subdivided into five sections for better read-
ability, representation and understanding of the readers. First-order and second-order
thermal modeling of LIB and the respective temperature estimation strategy are presented
in Section 2. The experimental setup and model parameter identification are discussed
in Section 3. Temperature estimation using the fusion of the first-order thermal model
with KF and second-order thermal model with KF is described in Section 4. Section 4 also
includes the comparative study between the first-order thermal model and second-order
thermal model in terms of prediction accuracy and modeling complexity. Major findings
and concluding remarks are drawn in the conclusion in Section 5.
2. Thermal Modeling and Temperature Estimation Strategy

Commercially, LIBs are available in many different form factors such as prismatic
cells, pouch cells [49] and cylindrical cells. Among these, cylindrical cells are widely
used in large-scale high-power applications. However, the cylindrical cell has worse
thermal heat dissipation, and the spiral format leads to a big thermal gradient inside
the cell. Therefore, the thermal modeling of a 18,650 cylindrical LIB cell is considered
in this study, considering the necessity of effective thermal management of cylindrical
LIB. The mathematical analysis and the fusion of KF with these thermal models for core
and surface temperature estimation are presented in this section. The aim is to provide a
guideline for selecting an appropriate thermal model for online prediction with an optimum
computational cost suitable for onboard low-cost BMS. As previously discussed in the
introduction section, the temperature estimation model consists of one heat generation
21
Energies 2022, 15, 623
model and a heat transfer model, where the heat generation model provides input to the
heat transfer model. Therefore, the modeling strategy and mathematical analysis of the
ECM-based heat generation model are considered here as well.
2.1. Heat Generation Model

The Electric Circuit Model (ECM) [50]—based thermal estimation model has been
reported to estimate the total heat generation inside the LIB cell by several researchers.
So far, electrochemical modeling has demonstrated the best performance in capturing the
nonlinearities of LIB, while at the same time, it is the most complex to model. Capturing the
high degree of nonlinearities higher-order ECM is required; however, the computational
cost and modeling complexity increase with the increase in model order. Yet the major
advantage of ECM is that a balance between the modeling complexity and model accuracy
can be achieved through optimization with the help of the model order reduction tech-
nique [51,52]. Therefore, a 1-RC (first-order) ECM is considered here to quantify the total
heat generation. The 1-RC ECM of LIB is shown in Figure 1. The basic strategy used by any
ECM-based heat generation model is to mathematically accumulate the heat generation
from internal power losses that typically depend on the internal resistance and charging–
discharging current level. Again, the heat generation depends on the cell SOC, current
level and temperature, as the internal resistances are the functions of these variables.
Figure 1. 1-RC ECM (Thevenin’s equivalent) model of an LIB cell.
The VOCV and V in Figure 1 represent the open-circuit voltage and the terminal
voltage, respectively. The steady-state DC series resistance, which represents the electrolyte
resistance to the lithium-ion transportation, is denoted as R0 in Figure 1. Further, the short
transient response is caused by the lithium-ion flow in the solid electrolyte interphase layer,
and the anode electrode is represented by polarization resistance (R1 ) and capacitance (C1 ),
respectively. These R1 and C1 appear only during the transient period [53]. A 1-RC battery
model was considered in this study due to its optimum performance, ease of modelling,
low computational cost and adequate accuracy when compared to other higher-order
RC models [54,55]. Further, the online determination of heat generation inside LIB with
these higher-order models is challenging due to computational cost. For this, Bernardi
et al. [56] developed a simplified equation for LIB heat generation calculationthatis suitable
for online prediction over other computationally expensive methods such as constant heat
generation rate [57], curve fitting technique [58] and Joule’s Law [59]-based methods. The
governing equation for the total heat generated inside the battery (Q) as developed by
Bernardi et al. [56] is shown in Equation (1).
Q = I (V − VOCV ) (1)
The parameters of this equation are also the function of charging–discharging current
(I), SOC and temperature, which are estimated using the ECM of the cell. Finally, the
22
Energies 2022, 15, 623
value of the Q, obtained from Equation (1), is used as one of the inputs to the first-order
and second-order thermal model for temperature estimation, which is discussed in the
following section.
2.2. First-Order Thermal Modelling

2.2.1. Mathematical Analysis of First-Order Thermal Model
Now, for the first-order model, as is noted by several other researchers, the surface
temperature is considered constant throughout the surface of the cell. Heat transport is only
along the radial direction, meaning the lateral surface temperature is considered the same
as axial direction (cell temperature at two terminals), as reported in [43]. Further, regarding
heat transfer, only heat conduction from the core to the surface is considered. Heat exchange
between surface and ambient by convection is not considered. The first-order thermal
model is depicted in Figure 2.
Figure 2. First-order heat resistor–capacitor-based thermal model of LIB.
In Figure 2, the thermal parameters, that is, the heat capacity of the core, heat transfer
resistance inside the cell, heat transfer resistance outside the cell and total quantity of heat
liberated concentrated from the core, are represented by Cc (J/K), Rc (K/W), Ru (K/W) and
Q (J), respectively. The unit of each respective quantity is mentioned in the parentheses. The
temperature of the core, surface and ambient is represented by Tc , Ts and Tamb , respectively,
measured in K. The core temperature at node Tc and surface temperature at node Ts can
be monitored using this model; thus this type of model is also referred to as a two-node
or two-state thermal model [22,60].The heat resistor–capacitor model uses the analogy
between the thermal and electrical systems, as discussed in the introduction section. Thus,
for mathematical analysis, the heat transfer rate is represented by electrical current (i), and
the branch currents are represented by ia , ib in the respective branch, as shown in Figure 2.
Therefore, the governing equation of the model can be derived by applying Kirchhoff’s
Current Law (KCL) at the Tc node. The current balance equation at node Tc reads:
i = i a + ib = Q (2)
Now, by rewriting Equation (2) in terms of thermal parameters, Equation (3) can
be found:
dTc Ts − Tc T − Ts
Q = Cc + + amb (3)
dt Ru Rc
By re-arranging Equation (3) we find:
dTc Ts − Tc T − Ts
Cc = Q+ + amb (4)
dt Ru Rc
23
Energies 2022, 15, 623
Finally, the value of Tc can be calculated by integrating Equation (4) with respect to
the total heat transfer time while the values of Ts and Tamb are known. While Tamb can
be easily measured by employing only one temperature sensor, the measurement of Ts
with physical sensors in a high-power LIB pack is challenging. Therefore, the alternative
solution is to estimate the surface temperature using a temperature estimation scheme.
One such estimation scheme is also proposed in reference [8], which estimates Ts from
known Tc .
2.2.2. KF for First-Order Thermal Model

KF is used to estimate and predict an unknown parameter from known parameters.
The state model for a KF and the first-order model, as developed in the reference [43,61]
and in [62] respectively, are also considered for this study. Now, assuming the state as Tc,t ,
output as Ts,t and inputs as Q and Tamb , The state-space matrices are derived by linearizing
Equation (4) in the discrete domain. A linearized version of Equation (4) is shown in
Equation (5).
Q Ts,t−1 − Tc,t−1 T − Ts,t−1
Tc,t − Tc,t−1 = t−1 + + amb,t (5)
Cc Cc Rc Cc Ru
As shown in reference [8], small changes in Ts can be ignored. Hence, the term Ts,t−1
can be considered as zero.
Q t −1 1 T
Tc,t = + Tc,t−1 (1 − ) + amb,t−1 (6)
Cc Cc Rc Ru Cc
The transfer matrices of the KF-based temperature estimation model can be found by
reducing Equation (6) in the form of state models as shown in Equations (7)–(9).
Hence,
1
A = [1 − ] (7)
Cc Rc
1 1
B=[ ] (8)
Cc Cc Ru
C=D=0 (9)
2.3. Second-Order Thermal Modelling

2.3.1. Mathematical Analysis of Second-Order Thermal Model
The condition of non-uniform Ts and heat transport in the radial direction through
conduction from the core to surface is also considered during the second-order thermal
modeling. Additionally, the heat exchange between the surface and ambient is considered in
the second-order model, which was not included in the first-order model. Only convective
heat exchange between the cell surface and ambient is considered here. Therefore, in
addition to the thermal properties of the first-order model, the thermal capacitance of cell
case (Cs ) is also considered. The resulting equivalent circuit of the second-order thermal
model using heat resistor–capacitor is shown in Figure 3, similarly to the findings of other
studies [8,60,63,64].
24
Energies 2022, 15, 623
Figure 3. Second-order equivalent circuit thermal model of LIB.
Q needs to be estimated for the same ECM-based strategy mentioned in Section 2.1. To
derive the mathematical analysis of the second-order thermal model, heat balance analysis
in the core and surface is performed. The heat balance equation at the core and surface is
represented in Equations (10) and Equation (11), respectively [8].
dTc Ts − Tc
Cc = Q+ (10)
dt Rc
dTs ( T − Ts ) ( Ts − Tc )
Cs = amb − (11)
dt Ru Rc
2.3.2. KF for Second-Order Thermal Model

Tc could be estimated by re-arranging the coupled ordinary differential equations
of the second-order thermal model. Since the thermal model has two thermal energy
storage parameters (Cc and Cs ), two governing equations are used to estimate Tc in terms
of measured Ts and Tamb .
1
A = [1 − ] (12)
Cc ( Rc + Ru )
1 1
B=[ ] (13)
Cc Cc ( Rc + Ru )
Ru
C=[ ] (14)
Rc + Ru
Rc
D = [0 ] (15)
Rc + Ru
It is worth noting that Cc , Rc and Ru in the second-order thermal model are the same
as Cp , Rin and Rout , respectively, in the first-order model.
2.3.3. Fundamentals of KF
It is worth providing a basic explanation of KF as it is the heart of the temperature
estimation scheme discussed here. A KF is a linear quadratic estimator and is mainly used
in statistics and control engineering. It outputs the estimates of an unknown state and uses
the noise and the inaccuracies of the measured output. Some of the common examples of
KF usage include guidance, navigation and core temperature estimation in EVs; the general
form of KF is shown below:
Xk = Ak−1 Xk−1 + Bk−1 Uk−1 + Wk−1 (16)
25
Energies 2022, 15, 623
Yk = Ck Xk + Dk Uk + Vk (17)
where Xk is the state of the system (Tc,t ), Yk is the output of the system (Ts,t ), Uk is the input
to the system ([Tamb,t Q]T ), t presents the state of the system and t−1 represents the previous
state of the system. The block diagram of a KF is shown in Figure 4. It is a robust and simple
technique used to estimate data based on its input signal. It uses mathematical modeling
of the system and by giving the same input as an actual system, it predicts the output.
The measured output from the actual system and predicted output from the mathematical
model are then compared to obtain the error. This error is multiplied with Kalman gain
and is added to the predicted state to obtain an accurate estimated state [65].
Figure 4. Fundamental building blocks of KF based estimation scheme.
3. Experimental Analysis for Thermal Model Parameterization

An automated battery testing system is the best option to collect battery test data, espe-
cially for an LIB, as LIB cells are highly sensitive to voltage, current, temperature and other
environmental uncertainties. Therefore, an in-house “Battery Automated System (BAS)”
was previously developed by the research group of Smart Transportation Electrification and
Energy Research (STEER). The setup was used to invent the constant temperature constant
current (CT-CV) charging technique [66,67] and several other prominent research studies
in the BMS domain [2,4,8]. The BAS is an experimental setup with a fully programmable
test environment control and data acquisition system. A schematic layout of the BAS is
shown in Figure 5. The experimental data were then used for the parameter estimation of
ECM and thermal modeling, model validation and model-to-model comparison purposes.
Interested readers are invited to refer to these papers [2,4,8,66,67] for more details about
the experimental setup. However, a brief overview of the experimental setup and test
conditions is also mentioned in this section as a quick reference for the readers. The basic
idea was to identify the input parameters of the thermal model, that is, heat capacity and
heat transfer coefficients, through a steady-state analysis as well as transient experiments
based on the nonlinear least square algorithm. The LIB cell was tested at three different
temperatures where the internal battery temperature was raised using standard current
pulses that were within the permissible limit specified on the manufacturer datasheet to
ensure no capacity fade occurred during testing.
26
Energies 2022, 15, 623
Figure 5. Schematic layout of the Battery Automated System (BAS).
Experimental Setup
Battery testing was performed on a 18,650 NMC (Lithium Nickel Manganese Cobalt
Oxide) LIB, manufactured by LG Chem. Detailed specifications of the cell as provided
by the manufacturer are mentioned in Table 1. A programmable power supply (Model:
E36313A) from Keysight and a programmable electronic load (Model: BK8601) from B&K
Precision were used for charging and discharging the battery with a predefined charging–
discharging current profile. Further, a programmable temperature chamber was used to
maintain the Tamb based on a predefined set-point. Finally, to control the BAS a MATLAB
script-based program was used. A programmable data acquisition system (DAQ) (model
DPM66204) from Chroma was used to collect the cell voltage, current and temperature data.
Different current profile sat three different ambient temperatures (Tamb = 273 K (0 ◦ C), 293 K
(20 ◦ C) and 323 K (50 ◦ C)) were used for charging and discharging experiments. Finally, a
nonlinear least square algorithm was used for online parameter estimation for developing
the ECM and thermal model as demonstrated by Surya et al. [8]. All the model components
were designed in MATLAB using three-dimension interpolated look up tables where the
feature vectors were SOC, Ibat and Tamb . The heat generation model and the first-order
and second-order thermal model were also developed in the MATLAB/Simulink and
Simscape environment. Finally, an extensive simulation study was conducted to collect the
simulated core and surface temperature data for further analysis. Simulation results were
used for model validation as well as model-to-model comparison between the first-order
and second-order thermal models. The core temperature (Tc ) was estimated using a KF
for various patterns of currents that were within the permissible limit specified on the
manufacturer datasheet to ensure no capacity fade occurred during testing.
27
Energies 2022, 15, 623
Table 1. Specifications of 18,650 LIB cell under test.
Specification Name Values

Manufacturer and Model LG Chem/INR18,650HG2
Cell Form Factor Cylindrical (18,650)
Chemistry Lithium Nickel Manganese Cobalt Oxide (LiNiMnCoO2 )
Nominal Voltage 3.6 V
Nominal Capacity 3 Ah
Standard Charging (CC-CV) 1.5 A, 4.2 V, Cut-off: 50 mA
Fast Charging (CC-CV) 4.0 A, 4.2 V, Cut-off: 100 mA
Discharging Condition 20 A (Max. Current), 2.5 V (Cut-off Voltage)
Operating Temperature Charge: 0 to 50 ◦ C, Discharge: −30 to 60 ◦ C
Pack Weight 48 g
4. Results and Discussion

This research study intended to answer whether it is worth developing a second-
order model instead of a first-order model for online temperature prediction by low-cost
onboard BMS, firstly, by developing a first-order and second-order thermal model utilizing
battery test data and MATLAB-based online parameter estimation; secondly, by simulating
the temperature profile of the cell using the first-order and second-order thermal models
subjected to different current profiles. The intention was to investigate the impact of
charging–discharging current on the core and surface temperature of the cell. Thirdly, we
compared the estimation results obtained from the first-order and second-order models. All
simulations were carried out in the MATLAB Simulink environment, where a fixed solver
and an appropriate step time were used [62]. Initially, simulation was carried out without
employing a KF to deduce the baseline analysis. Figure 6 shows the current profile used
for the base case analysis, and Figure 7 shows the plots of estimated Tc , Ts and the measure
Tamb . Previously, we measured Ts from experiments. By comparing the measured and
estimated Ts it was observed that both Ts were within the acceptable limit, and Tc and Ts
closely followed the current profile, and Tc > Ts > Tamb , as per the expectation, confirming
the modeling accuracy.
Figure 6. The pattern of the discharging current applied to the cell.
28
Energies 2022, 15, 623
Figure 7. The plot of Tc , Ts and Tamb without using KF.
In the subsequent sections, firstly, Tc was estimated using the combined first-order
thermal model and KF for three different current profiles and ambient temperatures, which
are illustrated in Case 1, Case 2 and Case 3, respectively. Secondly, a similar study was
also conducted for the second-order model and finally, the results were compared. All
experiments were carried out with different current profiles as per the manufacturer’s
recommendation to ensure no battery health degradation [58]. In all cases, the initial
currents were kept high for rapid charging of the cell.
4.1. Case 1: Tamb = 293 K (20 ◦ C)

At first, Tc was initialized to Ts in the simulation as initially, the cell was in a thermal
equilibrium state. Tamb was considered as 293 K (20 ◦ C), and a very low value of discharging
current was applied for the core and surface temperature to rise. The pattern used in Case
1 is shown in Figure 8, and the plot of estimated Tc and measured Ts are shown in Figure 9
whereas the difference between the estimated Tc and measured Ts is shown in Figure 10.
It was observed that Tc and Ts closely followed the current pattern, and the maximum
difference between estimated Tc and measured Ts was noted as 6.8 K, whilst it was also
noticed that for the entire duration, Tc > Ts, and the maximum difference occurred when
the current was at its peak.
29
Energies 2022, 15, 623
Figure 9. The plot of the estimated Tc and measured Ts .
Figure 10. Variation of the difference between the estimated Tc and measured Ts .
4.2. Case 2: Tamb = 323 K (50 ◦ C)

In the second phase of the experiments, the temperature of the thermal chamber (Tamb )
was set to 323 K (50 ◦ C). The pattern used in Case 2 is shown in Figure 11. Similar to Case 1,
Tc was initialized to Ts during the simulation here as well. The estimated Tc and measured
Ts are shown in Figure 12. It was observed that the temperatures closely followed the
current pattern here also. The maximum difference between Tc and measured Ts was noted
as 7K. The plot of the difference between the estimated Tc and measured Ts is shown in
Figure 13. Similar observations to those made for Case 1 were also noticed here in Case 2
regarding Tc and Ts .
30
Energies 2022, 15, 623
4.3. Case 3: Tamb =273 K (0 ◦ C)

During Case 3, the temperature of the thermal chamber (Tamb ) was set to 273 K (0 ◦ C),
and Tc was set equal to Ts . The pattern of discharging current applied to the battery is
shown in Figure 14. Figure 15 shows the estimated Tc and measured Ts . Figure 16 shows the
31
Energies 2022, 15, 623
difference between the estimated Tc and measured Ts . It can be noticed from Figure 15 that
at the beginning the magnitude of Tc and Ts were very large. This was due to the high value
of discharging current during this period. It was also observed that the temperature rise
is a slow process due to the presence of thermal resistances (Ru and Rc ). The temperature
difference increased as the value of discharge current increased. Therefore, it can be inferred
from these observations that the temperature rise closely follows the current through the
battery, and the rate of rising of Tc was the same as Ts for a low value of current. However,
for higher values of the current the rise in Tc was much higher than that in Ts . From these
observations, the importance of accurate core and surface temperature estimation alongside
the requirement of effective and efficient thermal management to maintain Tc under the
safe operating limit is evidenced.
32
Energies 2022, 15, 623
4.4. Comparison between First-Order and Second-Order Thermal Models

This section deals with the comparative analysis between the first-order and second-
order thermal based on the estimation accuracy, parameter identification, experimental test
requirement and suitability for onboard low-cost BMS. To compare the models, the same
values of thermal parameters, current, Ts , Tamb and Q were injected into the thermal models.
Similar current profiles to those used in Case 1, Case 2 and Case 3 of the first-order model
were also applied to the second-order thermal model. Heat generation was calculated
using the same 1-RC ECM as used in the first-order model. Finally, the estimated Tc profiles
obtained from the first-order and second-order thermal models were compared to analyze
the prediction accuracy of these models. Figure 17 shows the current profile used for the
comparative study, whereas Figures 18 and 19 depict the difference in Tc and Ts obtained
from the first-order and second-order thermal model, respectively.
Figure 17. The pattern of the discharging current applied to both the models.
33
Energies 2022, 15, 623
Figure 18. Difference between Tc and Ts obtained from the second-order thermal model.
Figure 19. Difference between Tc and Ts obtained from the first-order thermal model.
It was observed that the difference in temperatures was larger in the first-order ther-
mal model due to the change in Tc and not Ts . This is because of the decoupling between
Ts and Tc , as seen in Equation (3). Moreover, while comparing Equations (7)–(9) with
Equations (12)–(14), it was noticed that the output parameter Ts in KF showed no depen-
dence on the state Tc,t −1 which is also a major reason behind the estimation error in case
of the first-order model. Further, references [43,61] demonstrated that Cc and Ru of the
second-order thermal model have a significant effect on Tc . Since these parameters were
not present in the C and D matrices of the first-order model, a large increase in Tc was
observed. The thermal parameter sensitivity analysis, as conducted in references [8,61],
also confirmed the same reason behind the difference in temperature estimation by the
first-order thermal model. It was found that the difference between Ts and Tc is increased
if the discharge current increases. Hence, for currents with dynamic changes, Tc estima-
tion using the first-order model provides a large difference from the second-order model.
Further, Cc only contributed to the transient part of Tc . However, with small changes in
Rc and Ru , a large variation in Tc was also observed. The modeling complexity, experi-
mental requirement and computational expenses in the used second-order model were not
considerably high compared to the first-order model considered here. A tradeoff between
34
Energies 2022, 15, 623
the modeling complexity and accuracy requirement suggests the implementation of a

second-order model is worthwhile for smart BMS, especially for high-power applications
of LIB.
4.5. Comparison between First-Order and Second-Order Thermal Models for Higher C Rates
As discussed in the introduction, the performance of different types of battery models
is highly influenced by the value of charging–discharging current. As was already wit-
nessed from the above discussion, the second-order model is more accurate compared to
the first-order model. However, it is equally important to assess the performance of the
second-order model in a high value of discharge current for almost all practical purposes a
high value of discharge current is used. Therefore, a discharge current of 5A was applied
to both the first and second-order thermal models to observe the change in Tc . and Ts . The
difference between the estimated Tc and estimated Ts for the first and second-order thermal
models is shown in Figure 20.
Figure 20. Comparison between Tc -Ts for higher C discharge.
It was observed that the error (Tc -Ts ) was higher in the first-order model than in the
second-order model. Therefore, it could be concluded that the second-order model can also
predict a highly accurate temperature state in practical applications as well.
5. Conclusions
This paper deals with the core temperature (Tc ) estimation of lithium-ion 18,650 cell
using a Kalman filter (KF). This estimation provides effective thermal management, state
estimations, operational safety and the longer useful life of LIB. Initially, a detailed discus-
sion regarding the importance of core and surface temperature estimation was presented
followed by a review of the state-of-the-art temperature estimation strategies and thermal
modeling of LIB. Equivalent Circuit Models (ECM) of LIB-based heat generation model
and heat resistor–capacitor-based thermal models were developed in a MATLAB/Simulink
environment. Regarding heat resistor–capacitor-based thermal modeling, one first-order
and one second-order thermal model were developed and validated using laboratory ex-
perimental data. Further, extensive simulation studies were conducted to demonstrate the
influence of battery current and ambient temperature on the core and surface temperature
of the LIB cell. The heat transfer equations for a first-order and second-order thermal model
were derived, modeled and simulated. KF with appropriate process and measurement
noise levels was also used to estimate Tc in terms of measured surface (Ts ) and ambient
temperature (Tamb ). Finally, these results were compared to assess the prediction accuracy
35
Energies 2022, 15, 623
of these models. The difference between the core and surface temperatures was noted as
approximately 7 K to 8 Kin the first-order model, whereas it was only about 1 K to 2 Kin
the second-order thermal model. Ts showed no dependence on Tc in the first-order thermal
model. Further, the output parameter Ts in KF showed no dependence on the state Tc,t −1 ,
which is also a major reason behind the estimation error in the case of the first-order model.
The thermal capacitance of core (Cc ) and resistances (Ru ) of the second-order thermal model
have a significant effect on Tc . Since these parameters are not present in the C and D matri-
ces of the first-order model, a large increase in Tc was observed in the first-order thermal
model. Hence, the inaccuracy was only due to the error in Tc estimation. The findings are
also supported by several other research studies in the domain. Further, the consideration
of the thermal capacitance of cell casing and the impact of ambient conditions on the second-
order model were the reasons for high accuracy. Further, the performances of first and
second-order thermals were also judged with a high value of discharge current for assessing
their performance during practical operation. It was observed that the second-order model
performance was highly satisfactory compared to the first-order model even in practical
applications typically requiring a high value of discharge current. However, estimating
the additional parameters of the second-order model requires more experimental data and
time. Moreover, due to the complex mathematical form of the second-order model, it takes
more computation time. However, looking at the prediction accuracy and the increasing
stringent requirement of highly accurate states of battery, it could be stated that it is worth
investing more time, cost and expertise in developing a second-order thermal model for
more accurate temperature estimation in LIB. This is especially true for the advanced BMS
required for high-power LIB packs used in EVs and grid-tied energy storage alongside
highly sophisticated consumer electronics. The discussed second-order thermal of a single
cell can be extended to an LIB pack by integrating the thermal gradient and the impact of
peripheral cells alongside optimal placing of temperature sensors inside the battery casing
to adjust the ambient temperature parameter value in the model. All these aspects will be
considered in our future research.
Author Contributions: Conceptualization, S.S. and S.W.; methodology, S.S. and A.S.; software, S.S.
and A.S.; validation, S.S.; formal analysis, S.S. and A.S.; investigation, S.S. and A.S.; resources, V.M.
and S.W.; data curation, V.M.; writing—original draft preparation, S.S. and A.S.; writing—review
and editing, A.S. and S.W.; visualization, S.S. and A.S.; supervision, S.W.; All authors have read and
agreed to the published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Conflicts of Interest: The authors declare no conflict of interest. No data/information from Robert
Bosch Engineering and Business Solutions Private Limited (RBEI) were used for this work.
References
1. Kleiner, J.; Stuckenberger, M.; Komsiyska, L.; Endisch, C. Advanced Monitoring and Prediction of the Thermal State of Intelligent
Battery Cells in Electric Vehicles by Physics-Based and Data-Driven Modeling. Batteries 2021, 7, 31. [CrossRef]
2. Marcis, V.A.; Kelkar, A.; Williamson, S.S. Electrical circuit modeling of a 18650 lithium-ion cell for charging protocol testing for
transportation electrification applications. In Proceedings of the 2020 IEEE Transportation Electrification Conference & Expo
(ITEC), Chicago, IL, USA, 23–26 June 2020.
3. Dao, V.Q.; Dinh, M.-C.; Kim, C.S.; Park, M.; Doh, C.-H.; Bae, J.H.; Lee, M.-K.; Liu, J.; Bai, Z. Design of an Effective State of Charge
Estimation Method for a Lithium-Ion Battery Pack Using Extended Kalman Filter and Artificial Neural Network. Energies 2021,
14, 2634. [CrossRef]
4. Marcis, V.A.; Praneeth, A.V.J.S.; Patnaik, L.; Williamson, S.S. Analysis of CT-CV Charging Technique for Lithium-ion and NCM
18650 Cells over Temperature Range. In Proceedings of the Proceedings of the IEEE International Conference on Industrial
Technology, Buenos Aires, Argentina, 26–28 February 2020.
36
Energies 2022, 15, 623
5. Li, W.; Zhu, J.; Xia, Y.; Gorji, M.B.; Wierzbicki, T. Data-Driven Safety Envelope of Lithium-Ion Batteries for Electric Vehicles. Joule
2019, 3, 2703–2715. [CrossRef]
6. Lisbona, D.; Snee, T. A review of hazards associated with primary lithium and lithium-ion batteries. Process Saf. Environ. Prot.
2011, 89, 434–442. [CrossRef]
7. Kong, L.; Li, C.; Jiang, J.; Pecht, M.G. Li-ion battery fire hazards and safety strategies. Energies 2018, 11, 2191. [CrossRef]
8. Surya, S.; Marcis, V.; Williamson, S. Core Temperature Estimation for a Lithium ion 18650 Cell. Energies 2020, 14, 87. [CrossRef]
9. Zhu, J.; Sun, Z.; Wei, X.; Dai, H. Battery internal temperature estimation for LiFePO4 battery based on impedance phase shift
under operating conditions. Energies 2017, 10, 60. [CrossRef]
10. Sidhu, M.S.; Ronanki, D.; Williamson, S. State of charge estimation of lithium-ion batteries using hybrid machine learning
technique. In Proceedings of the IECON Proceedings (Industrial Electronics Conference), Lisbon, Portugal, 14–17 October 2019.
11. Sidhu, M.S.; Ronanki, D.; Williamson, S. Hybrid State of Charge Estimation Approach for Lithium-ion Batteries using k-Nearest
Neighbour and Gaussian Filter-based Error Cancellation. IEEE Int. Symp. Ind. Electron. 2019, 2019, 1506–1511. [CrossRef]
12. Samanta, A.; Chowdhuri, S. Active Cell Balancing of Lithium-ion Battery Pack Using Dual DC-DC Converter and Auxiliary
Lead-acid Battery. J. Energy Storage 2021, 33, 102109. [CrossRef]
13. Kelkar, A.; Dasari, Y.; Williamson, S.S. A Comprehensive Review of Power Electronics Enabled Active Battery Cell Balancing for
Smart Energy Management. In Proceedings of the 2020 IEEE International Conference on Power Electronics, Smart Grid and
Renewable Energy (PESGRE2020), Cochin, India, 2–4 January 2020.
14. Samanta, A.; Chowdhuri, S.; Williamson, S.S. Machine Learning-Based Data-Driven Fault Detection/Diagnosis of Lithium-Ion
Battery: A Critical Review. Electronics 2021, 10, 1309. [CrossRef]
15. Tanim, T.R.; Rahn, C.D.; Wang, C.Y. State of charge estimation of a lithium ion cell based on a temperature dependent and
electrolyte enhanced single particle model. Energy 2015, 80, 731–739. [CrossRef]
16. Farmann, A.; Sauer, D.U. A study on the dependency of the open-circuit voltage on temperature and actual aging state of
lithium-ion batteries. J. Power Sources 2017, 347, 1–13. [CrossRef]
17. Zheng, F.; Jiang, J.; Sun, B.; Zhang, W.; Pecht, M. Temperature dependent power capability estimation of lithium-ion batteries for
hybrid electric vehicles. Energy 2016, 113, 64–75. [CrossRef]
18. Surya, S.; Mn, A. Effect of Fast Discharge of a Battery on its Core Temperature. In Proceedings of the 2020 International
Conference on Futuristic Technologies in Control Systems & Renewable Energy (ICFCR), Malappuram, India, 23–24 September
2020. [CrossRef]
19. Li, Z.; Zhang, J.; Wu, B.; Huang, J.; Nie, Z.; Sun, Y.; An, F.; Wu, N. Examining temporal and spatial variations of internal
temperature in large-format laminated battery with embedded thermocouples. J. Power Sources 2013, 241, 536–553. [CrossRef]
20. Robinson, J.B.; Darr, J.A.; Eastwood, D.S.; Hinds, G.; Lee, P.D.; Shearing, P.R.; Taiwo, O.O.; Brett, D.J.L. Non-uniform temperature
distribution in Li-ion batteries during discharge-A combined thermal imaging, X-ray micro-tomography and electrochemical
impedance approach. J. Power Sources 2014, 252, 51–57. [CrossRef]
21. Raijmakers, L.H.J.; Danilov, D.L.; Eichel, R.A.; Notten, P.H.L. A review on various temperature-indication methods for Li-ion
batteries. Appl. Energy 2019, 240, 918–945. [CrossRef]
22. Pan, Y.W.; Hua, Y.; Zhou, S.; He, R.; Zhang, Y.; Yang, S.; Liu, X.; Lian, Y.; Yan, X.; Wu, B. A computational multi-node
electro-thermal model for large prismatic lithium-ion batteries. J. Power Sources 2020, 459, 228070. [CrossRef]
23. Sun, F.; Xiong, R.; He, H.; Li, W.; Aussems, J.E.E. Model-based dynamic multi-parameter method for peak power estimation of
lithium-ion batteries. Appl. Energy 2012, 96, 378–386. [CrossRef]
24. Lu, L.; Han, X.; Li, J.; Hua, J.; Ouyang, M. A review on the key issues for lithium-ion battery management in electric vehicles. J.
Power Sources 2013, 226, 272–288. [CrossRef]
25. Ghalkhani, M.; Bahiraei, F.; Nazri, G.A.; Saif, M. Electrochemical–Thermal Model of Pouch-type Lithium-ion Batteries. Electrochim.
Acta 2017, 247, 569–587. [CrossRef]
26. Yang, X.G.; Leng, Y.; Zhang, G.; Ge, S.; Wang, C.Y. Modeling of lithium plating induced aging of lithium-ion batteries: Transition
from linear to nonlinear aging. J. Power Sources 2017, 60, 28–40. [CrossRef]
27. Allafi, W.; Zhang, C.; Uddin, K.; Worwood, D.; Dinh, T.Q.; Ormeno, P.A.; Li, K.; Marco, J. A lumped thermal model of lithium-ion
battery cells considering radiative heat transfer. Appl. Therm. Eng. 2018, 143, 472–481. [CrossRef]
28. Esmaeili, J.; Jannesari, H. Developing heat source term including heat generation at rest condition for Lithium-ion battery pack by
up scaling information from cell scale. Energy Convers. Manag. 2017, 139, 194–205. [CrossRef]
29. Arora, S.; Shen, W.; Kapoor, A. Neural network based computational model for estimation of heat generation in LiFePO4 pouch
cells of different nominal capacities. Comput. Chem. Eng. 2017, 101, 81–94. [CrossRef]
30. Chen, M.; Bai, F.; Song, W.; Lv, J.; Lin, S.; Feng, Z.; Li, Y.; Ding, Y. A multilayer electro-thermal model of pouch battery during
normal discharge and internal short circuit process. Appl. Therm. Eng. 2017, 120, 506–516. [CrossRef]
31. Zhao, Y.; Diaz, L.B.; Patel, Y.; Zhang, T.; Offer, G.J. How to Cool Lithium Ion Batteries: Optimising Cell Design using a Thermally
Coupled Model. J. Electrochem. Soc. 2019, 166, A2849–A2859. [CrossRef]
32. Damay, N.; Forgez, C.; Bichat, M.P.; Friedrich, G. Thermal modeling of large prismatic LiFePO4/graphite battery. Coupled
thermal and heat generation models for characterization and simulation. J. Power Sources 2015, 283, 37–45. [CrossRef]
33. Hu, X.; Liu, W.; Lin, X.; Xie, Y. A Comparative Study of Control-Oriented Thermal Models for Cylindrical Li-Ion Batteries. IEEE
Trans. Transp. Electrif. 2019, 5, 1237–1253. [CrossRef]
37
Energies 2022, 15, 623
34. Xie, Y.; Li, W.; Hu, X.; Zou, C.; Feng, F.; Tang, X. Novel Mesoscale Electrothermal Modeling for Lithium-Ion Batteries. IEEE Trans.
Power Electron. 2020, 35, 2595–2614. [CrossRef]
35. Friesen, A.; Mönnighoff, X.; Börner, M.; Haetge, J.; Schappacher, F.M.; Winter, M. Influence of temperature on the aging behavior
of 18650-type lithium ion cells: A comprehensive approach combining electrochemical characterization and post-mortem analysis.
J. Power Sources 2017, 342, 88–97. [CrossRef]
36. Fan, Y.; Bao, Y.; Ling, C.; Chu, Y.; Tan, X.; Yang, S. Experimental study on the thermal management performance of air cooling for
high energy density cylindrical lithium-ion batteries. Appl. Therm. Eng. 2019, 155, 96–109. [CrossRef]
37. Saw, L.H.; Poon, H.M.; Thiam, H.S.; Cai, Z.; Chong, W.T.; Pambudi, N.A.; King, Y.J. Novel thermal management system using
mist cooling for lithium-ion battery packs. Appl. Energy 2018, 223, 146–158. [CrossRef]
38. Liu, B.; Yin, S.; Xu, J. Integrated computation model of lithium-ion battery subject to nail penetration. Appl. Energy 2016, 183,
278–289. [CrossRef]
39. Doyle, M. Modeling of Galvanostatic Charge and Discharge of the Lithium/Polymer/Insertion Cell. J. Electrochem. Soc. 1993, 140,
1526–1533. [CrossRef]
40. Xiao, Y.; Fahimi, B. State-space based multi-nodes thermal model for lithium-ion battery. In Proceedings of the 2014 IEEE
Transportation Electrification Conference and Expo: Components, Systems, and Power Electronics-From Technology to Business
and Public Policy (ITEC 2014), Dearborn, MI, USA, 15–18 June 2014.
41. Tian, N.; Fang, H.; Wang, Y. 3-D Temperature Field Reconstruction for a Lithium-Ion Battery Pack: A Distributed Kalman Filtering
Approach. IEEE Trans. Control Syst. Technol. 2019, 27, 847–854. [CrossRef]
42. Ruan, H.; Jiang, J.; Sun, B.; Gao, W.; Wang, L.; Zhang, W. Online estimation of thermal parameters based on a reduced
wide-temperature-range electro-thermal coupled model for lithium-ion batteries. J. Power Sources 2018, 396, 715–724. [CrossRef]
43. Forgez, C.; Vinh Do, D.; Friedrich, G.; Morcrette, M.; Delacourt, C. Thermal modeling of a cylindrical LiFePO4/graphite
lithium-ion battery. J. Power Sources 2010, 195, 2961–2968. [CrossRef]
44. Dees, D.W.; Battaglia, V.S.; Bélanger, A. Electrochemical modeling of lithium polymer batteries. J. Power Sources 2002, 110, 310–320.
[CrossRef]
45. Lin, X.; Perez, H.E.; Siegel, J.B.; Stefanopoulou, A.G.; Li, Y.; Anderson, R.D.; Ding, Y.; Castanier, M.P. Online parameterization of
lumped thermal dynamics in cylindrical lithium ion batteries for core temperature estimation and health monitoring. IEEE Trans.
Control Syst. Technol. 2013, 21, 1745–1755. [CrossRef]
46. Choi, J.W.; Aurbach, D. Promise and reality of post-lithium-ion batteries with high energy densities. Nat. Rev. Mater. 2016, 1,
16013. [CrossRef]
47. Yang, F.; Wang, D.; Zhao, Y.; Tsui, K.L.; Bae, S.J. A study of the relationship between coulombic efficiency and capacity degradation
of commercial lithium-ion batteries. Energy 2018, 145, 486–495. [CrossRef]
48. Wang, Q.; Zhao, X.; Ye, J.; Sun, Q.; Ping, P.; Sun, J. Thermal response of lithium-ion battery during charging and discharging
under adiabatic conditions. J. Therm. Anal. Calorim. 2016, 124, 417–428. [CrossRef]
49. An, S.J.; Li, J.; Daniel, C.; Kalnaus, S.; Wood, D.L. Design and Demonstration of Three-Electrode Pouch Cells for Lithium-Ion
Batteries. J. Electrochem. Soc. 2017, 164, A1755–A1764. [CrossRef]
50. Patnaik, L.; Williamson, S. A Five-Parameter Analytical Curvefit Model for Open-Circuit Voltage Variation with State-of-Charge
of a Rechargeable Battery. In Proceedings of the 2018 IEEE International Conference on Power Electronics, Drives and Energy
Systems (PEDES 2018), Chennai, India, 18–21 December 2018.
51. Tran, N.T.; Farrell, T.; Vilathgamuwa, M.; Choi, S.S.; Li, Y. A Computationally Efficient Coupled Electrochemical-Thermal Model
for Large Format Cylindrical Lithium Ion Batteries. J. Electrochem. Soc. 2019, 166, A3059–A3071. [CrossRef]
52. Subramaniam, A.; Kolluri, S.; Santhanagopalan, S.; Subramanian, V.R. An Efficient Electrochemical-Thermal Tanks-in-Series
Model for Lithium-Ion Batteries. J. Electrochem. Soc. 2020, 167, 113506. [CrossRef]
53. Bryden, T.S.; Dimitrov, B.; Hilton, G.; Ponce de León, C.; Bugryniec, P.; Brown, S.; Cumming, D.; Cruden, A. Methodology to
determine the heat capacity of lithium-ion cells. J. Power Sources 2018, 395, 369–378. [CrossRef]
54. Lin, C.; Yu, Q.; Xiong, R.; Wang, L.Y. A study on the impact of open circuit voltage tests on state of charge estimation for
55. Surya, S.; Channegowda, J.; Datar, S.D.; Jha, A.S.; Victor, A. Accurate battery modeling based on pulse charging using MAT-
LAB/Simulink. In Proceedings of the 9th IEEE International Conference on Power Electronics, Drives and Energy Systems
(PEDES 2020), Jaipur, India, 16–19 December 2020.
56. Bernardi, D.; Pawlikowski, E.; Newman, J. A general energy balance for battery systems. J. Electrochem. Soc. 1985, 132, 5.
[CrossRef]
57. Greco, A.; Cao, D.; Jiang, X.; Yang, H. A theoretical and computational study of lithium-ion battery thermal management for
electric vehicles using heat pipes. J. Power Sources 2014, 257, 344–355. [CrossRef]
58. Rao, Z.; Wang, S.; Zhang, G. Simulation and experiment of thermal energy management with phase change material for ageing
LiFePO4 power battery. Energy Convers. Manag. 2011, 52, 3408–3414. [CrossRef]
59. Xie, Y.; Shi, S.; Tang, J.; Wu, H.; Yu, J. Experimental and analytical study on heat generation characteristics of a lithium-ion power
battery. Int. J. Heat Mass Transf. 2018, 122, 884–894. [CrossRef]
60. Chen, L.; Hu, M.; Cao, K.; Li, S.; Su, Z.; Jin, G.; Fu, C. Core temperature estimation based on electro-thermal model of lithium-ion
batteries. Int. J. Energy Res. 2020, 44, 5320–5333. [CrossRef]
38
Energies 2022, 15, 623
61. Surya, S.; Bhesaniya, A.; Gogate, A.; Ankur, R.; Patil, V. Development of thermal model for estimation of core temperature of
batteries. Int. J. Emerg. Electr. Power Syst. 2020, 21, 20200070. [CrossRef]
62. Surya, S.; Patil, V. Cuk Converter as an Efficient Driver for LED. In Proceedings of the 4th International Conference on Electrical,
Electronics, Communication, Computer Technologies and Optimization Techniques (ICEECCOT 2019), Mysuru, India, 13–14
December 2019.
63. Lin, X.; Stefanopoulou, A.G.; Perez, H.E.; Siegel, J.B.; Li, Y.; Anderson, R.D. Quadruple adaptive observer of the core temperature
in cylindrical Li-ion batteries and their health monitoring. Proc. Am. Control Conf. 2012, 578, 578–583. [CrossRef]
64. Sun, J.; Wei, G.; Pei, L.; Lu, R.; Song, K.; Wu, C.; Zhu, C. Online internal temperature estimation for lithium-ion batteries based on
Kalman filter. Energies 2015, 8, 4400–4415. [CrossRef]
65. Plett, G. BatteryManagement Systems, Volume II: Equivalent-Circuit Methods; Artech: Morristown, NJ, USA, 2015; ISBN 9781630810283.
66. Patnaik, L.; Praneeth, A.V.J.S.; Williamson, S.S. A Closed-Loop Constant-Temperature Constant-Voltage Charging Technique to
Reduce Charge Time of Lithium-Ion Batteries. IEEE Trans. Ind. Electron. 2019, 66, 1059–1067. [CrossRef]
67. Williamson, S.; Vincent, D.; Praneeth, A.; Sang, P.H. Charging Strategies for Electrified Transport. In Advances in Carbon
Management Technologies; CRC Press: Boca Raton, FL, USA, 2021; pp. 284–302.
39
energies
Article
Controlled Energy Flow in Z-Source Inverters
Zbigniew Rymarski, Krzysztof Bernacki * and Łukasz Dyga
Department of Electronics, Electrical Engineering and Microelectronics, Faculty of Automatic Control,

Electronics and Computer Science, Silesian University of Technology, Akademicka 16, 44-100 Gliwice, Poland;
zbigniew.rymarski@polsl.pl (Z.R.); lukasz.dyga@polsl.pl (Ł.D.)
* Correspondence: krzysztof.bernacki@polsl.pl
Abstract: This paper proposes a method to reduce the output voltage distortions in voltage source
inverters (VSI) working with impedance networks. The three main reasons for the voltage distortions
include a discontinuous current in the coils of the impedance network, the double output frequency
harmonics in the VSI’s voltage output caused by insufficient capacitance in the impedance network,
and voltage drops on the bridge switches during the shoot-through time. The first of these distortions
can be reduced by increasing the current of the impedance network when the output VSI current is
low. This method requires storing energy in the battery connected to the DC link of the VSI during the
“non-shoot through” time. Furthermore, this solution can also be used when the Z-source inverter
works with a photovoltaic cell to help it attain a maximum power point. The Z-source inverter
is essentially a voltage source inverter with the Z-source in the input. In this paper, the theory
behind basic impedance networks of Z-source and quasi-Z-source (qZ-source) is investigated where
simulations of the presented solutions and experimental verification of the results are also presented.
Keywords: impedance network; Z-source; quasi-Z-source; voltage source inverter; voltage distortions
Citation: Rymarski, Z.; Bernacki, K.;

Dyga, Ł. Controlled Energy Flow in
1. Introduction
Z-Source Inverters. Energies 2021, 14,
7272. https://doi.org/10.3390/ The Z-source impedance network was proposed initially by Peng [1]. This type of
en14217272 DC/DC converter was increasing the input DC voltage that is connected to a single-phase
or three-phase bridge voltage source inverter (VSI) which switches were used to store
Academic Editors: Sheldon energy in the coils of a Z-source. During shoot-through time, energy is stored when both
Williamson and Andrei Blinov switches in one of the inverter bridge legs are activated. This is only possible only in zero
states of the inverter. The modulation index M is restricted to the equation M = 1 − dZ
Received: 6 October 2021 where dZ = TST /Ts . The parameters TST , Ts , and dZ represent the shoot-through time,
Accepted: 1 November 2021 switching period of the inverter, and shoot-through time coefficient, respectively.
Published: 3 November 2021 For a Z-source, it is essential that the shoot-through time, dZ is less than 0.5. A voltage
source inverter with a Z-source is known as the Z-source inverter (ZSI). An impedance
Publisher’s Note: MDPI stays neutral network can function simply as a DC/DC converter with one additional switch in its
with regard to jurisdictional claims in output realizing shoot-through time but without an inverter. The input current of the Z-
published maps and institutional affil-
source is discontinuous (discontinuous input current—DIC) so Peng showed the changed
iations.
structure of the impedance network [2,3]. When a diode usually connected in series with
the input is replaced, this structure is called a qZ-source. As a result of this modification,
the new quasi-Z-source inverter (qZSI) structure is characterized by a continuous input
current (CIC) which has improved the use of an impedance network in photovoltaic
Copyright: © 2021 by the authors. (PV) systems [4]. Various methods of improving impedance networks structures have
Licensee MDPI, Basel, Switzerland. been developed [5] and a suitable example is the switched inductor Z-source inverter
This article is an open access article (SLZSI) [6]. The benefit of using these improved converters is a higher boost factor of the
input DC voltage than in the qZSI. Other existing impedance network structures include the
embedded SLZSI [7], an inductor-capacitor-capacitor-transformer Z-source (LCCTZSI) [8,9],
and a cascaded quasi-Z-source (CqZSI) [10]. The two-winding magnetically coupled
impedance source (MCIS) impedance network with a continuous input current [11] has a
4.0/).

Energies 2021, 14, 7272
high boost factor. The impedance network circuit based on three coupled inductors with
a delta (Δ) connection is presented in [12] and further developed in [13]. The networks
found in references [11] and [12] respectively were functional where an additional switch
was used without an inverter. A broad review of the impedance network topologies
is presented in [14,15], amongst other newly developed solutions based on impedance
networks [16–20]. Additionally, several methods of controlling impedance networks have
been considered which can be reviewed in [21,22]. However, the symmetric structure of a
Z-source with discontinuous input current due to a diode connected in series (Figure 1),
and an asymmetric quasi-Z-source (Figure 2) with maximum boost control is sufficient
to show the influence of an impedance network on VSI output voltage distortions and
proposed ways of reducing these distortions.
(a)
(b)
Figure 1. (a) Non-shoot-through state and (b) shoot-through state of the Z-source impedance network
with the VSI.
Further investigation of these improved network structures has shown that the power
efficiency of these systems including the decreased efficiency of the inverter is lower than
the efficiency of basic structures. Owing to this decreased efficiency the real boost factor
is also much lower than expected [23]. It is worth mentioning that significant differences
in recorded levels of radiated disturbances can be expected depending on the type of
impedance network structure used [24]. Unfortunately, additional losses in the switches
of the VSI during the shoot-through time are observed when switches are absent in the
impedance networks. Comparing the performance of a boost converter [23,25], it can be
shown that the VSI with an input synchronous boost converter can have a higher efficiency
than the same inverter with an impedance network.
42
Energies 2021, 14, 7272
(a)
(b)
Figure 2. (a) Non-shoot-through state and (b) shoot-through state of the qZ-source impedance
network with the VSI.
The basic structures of Z-source and qZ-source impedance networks are utilized today
in photovoltaic systems [26]. The main disadvantage of these impedance networks lies in
the discontinuous current mode (DCM) where the current in the inductors is equal to zero
for a time period during Ts where there is a low load of the VSI and a low dZ coefficient.
This is the main reason for the VSI output voltage distortions as shown in Figure 3a,b. By
calculating a sufficiently large inductance of the coils [23,27,28] and selecting an appropriate
magnetic material [29] for the lowest load while assuming the value of dZ , the current
in the coils should not decrease to zero. During operation, it cannot be guaranteed that
the load current will be nominal. Thus, the additional current taken from the impedance
network is a solution of DCM omitting for a low load current.
Another reason for output distortions is the insufficient capacity of Z-source capacitors.
Input current from a VSI bridge is like a “rectified” waveform that is filtered by the
LC input network and is approximately the first harmonic of the “rectified” current at
100 Hz. This means that 100 Hz distortion is present in the 50 Hz output waveform
as shown in Figure 3c. For the insufficient capacity, the output sinusoidal waveform is
left-skewed [23,27]. The third type of VSI output distortions are observed after crossing
zero output voltage caused by the additional voltage drops on the switched-on transistors
during the shoot-through time (see Figure 3a–c), thus causing oscillations after a change
of polarization in the PWM voltage. The impedance network influences the dynamic
properties of an entire ZSI [23,27,28] which introduces additional resonant frequencies and
the additional damping to the Bode plots of the ZSI. The main objective of this paper is
to demonstrate how charging the battery from a DC-link after the impedance network
during the non-shoot through times can reduce output distortions caused by the DCM
of the impedance network. However, charging a battery with too high a current can
lead to distortions of the output voltage after the voltage current is zero crossing and
oscillations as the result of the higher voltage drops on the switches during the shoot-
through time. Experimental results presented will show how charging the battery for a
Z-source decreases the output of total harmonic distortions (THD) even in the case when a
sophisticated feedback loop, for example, a passivity-based control (PBC), is used.
43
Energies 2021, 14, 7272
Figure 3. Inverter output voltage distortions, (a) Z-source in DCM using a VSI output filter capacitor
CF = 1 μF, (b) Z-source in DCM using a VSI output filter capacitor CF = 50 μF, (c) 100 Hz harmonic
distortions with a Z-source capacitor CZ = 100 μF.
Figure 3 presents the different types of output voltage distortions of the ZSI. In
Figure 3a,b, the DCM of the Z-source uses a low load current and ZSI output filter capaci-
tors of CF = 1 μF and 50 μF respectively. Figure 3c shows the distortions caused by a 100 Hz
current harmonic using a high load current and a Z-source capacitor of CZ = 100 μF.
Section 2 presents the basic structures of impedance networks and calculations of the
minimum ZSI output current IOUTrmsmin that ensure their continuous current mode (CCM).
In Section 3 the idea of the inverter with the impedance network charging the battery from
the DC link (during non-shoot-through time) to keep the impedance network in CCM
is presented. The simulations and results of the experimental verification are presented.
Section 4 contains the discussion of what kind of previously presented types of ZSI output
voltage distortions can be canceled by the controlled charging of the battery. Section 5
presents the final conclusions.
2. Basic Impedance Networks: Z-Source and qZ-Source

The Z-source and qZ-source impedance networks shown in Figures 1 and 2, respec-
tively, can operate in different states. Two basic states were taken into account during
analysis and these include the shoot-through and the non-shoot-through states. The non-
shoot-through state is depicted in Figures 1a and 2a, while the shoot-through state [23,27,28]
is shown in Figures 1b and 2b.
The Z-source has a symmetrical structure where the values of the inductors are equal
i.e., LZ1 = LZ2 . Similarly, the values of capacitors are the same, i.e., CZ1 = CZ2 , and the
currents in both inductors are the same, i.e., iLZ1 = iLZ2 . In the qZ-source, the currents
in both coils are the same and are identical to the Z-source coils currents (neglecting the
influence of the different parasitic resistances) if coils have equal inductances.
The amplitude of the VSI output voltage VOUTmax for the ZSI and qZSI is defined in
Equation (1) as
M
VOUTmax = ηkV MVDC = η V (1)
1 − 2d Z DC
44
Energies 2021, 14, 7272
where η is the efficiency, VDC is the input voltage, M is the VSI modulation coefficient, and
kV ’ is the DC voltage boost factor of the impedance network without power losses [23,27,28].
It is assumed that the capacitance CZ in the Z-source and qZ-source networks are suf-
ficiently high. The average voltage on the capacitors of the Z-source and the CZ2 capacitor
of the qZ-source are identical to the average voltage VLZav on the inductors [23,27,28] given
in Equation (2) as follows:
1 − dZ
VLZ1av = VLZ2av = VLZ = V (2)
1 − 2d Z DC
The input power PIN and output power POUT of the VSI connected to the impedance
networks for a Z-source or qZ-source can be calculated using Equations (3)–(5):
PI N = VDC IDCav = VDC ILZav (3)
POUT = VOUTrms IOUTrms = ηPI N (4)

1 M
POUT = √ η V I = ηVDC ILZav (5)
2 1 − 2d Z DC OUTrms
where ILZav is a single inductor current averaged over the fundamental period Tm .
For the simplest case of the resistive ZSI load, RLOAD the output power can be defined
Equation (6) as
2
1 M 1
POUT = √ η VDC = ηVDC ILZav (6)
2 1 − 2d Z R LOAD
And the average inductor current ILZav for the root mean square (rms) value of the
inverter output current IOUTrms is given Equation (7) as
1 M
ILZav = √ IOUTrms (7)
2 1 − 2d Z
The iLZ inductor current illustrated in Figure 4a comprises three components. These
components are the average current ILZav , the current iLZ2fm which is averaged in the Ts
switching period, and the triangle component iLZΔ of the inductor current. The current
iLZ2fm has the double fundamental frequency caused by the envelope of the input current
of the VSI bridge in the non-shoot-through time while the triangle component inductor
current iLZΔ is caused by storing energy in the coil during the shoot-through time and
recovering energy in the rest of the switching period (in CCM). A plot of the VSI input
current is displayed in Figure 4b.
The inductor current iLZ is defined in Equation (8) as
i LZ (t) = ILZav + i LZ2 f m (t) + i LZΔ (t) (8)
Figure 4 shows plots of a Z-source or qZ-source impedance network coil current and
an inverter input current including shoot-through current pulses for cases of maximum
and close to zero crossing of the inverter output voltage (in CCM).
This most important harmonic component 2 fm of the VSI bridge input current flows
through the LZ CZ circuit of the impedance network as shown in Equation (9). It is assumed
that all power losses are within the impedance network including the power losses on the
VSI switches during the shoot-through time.

4 √
1
i LZh2 f m ( abs(i LOAD (t))) = 2IOUTrms cos(4π f m t) (9)
3π 1 − (4π f m )2 L Z CZ
45
Energies 2021, 14, 7272
(a)
(b)
Figure 4. A Z-source or qZ-source impedance network (a) coil current and (b) the VSI input current
including shoot-through current pulses (that do not supply inverter) in the case of wide (for the
maximum of the output inverter voltage) and short (close to zero crossing of the output inverter
voltage) inverter PWM pulses in the CCM.
The triangle component iLZΔ of the inductor current iLZ in the CCM is calculated
approximately with the assumption that a sufficiently low capacitor voltage ripple ΔVCZ is
approximately equal to 0 and VCZmax is nearly equal to VCZav for the shoot-through time.
The triangle component iLZΔ can thus be expressed in Equation (10) as
VCZav VCZav 1 1− d Z
i LZΔ (t) ≈ L Z t, i LZΔmax = L Z Tst = L Z 1−2d Z VDC d Z Ts , i LZΔmax =
√ 1 1− d Z (10)
2 LZ η M VOUTrms d Z Ts
Consequently, the inductor current can be defined Equation (11) as

4 √
1 M 1
i LZ (t) = [ + 2 cos(4π f m t) ] I + i LZΔ (t) (11)
2 1 − 2d Z 3π 1 − (4π f m ) L Z CZ OUTrms
2
46
Energies 2021, 14, 7272
The lowest value of the inductor current is calculated Equation (12) as

√
1 M 4 2 1
1
i LZmin (t) = [ − ] I − i (12)
2 1 − 2d Z 3π 1 − (4π f m ) L Z CZ OUTrms 2 LZΔmax
2
As shown in Figure 4a, the requirement for CCM is that iLZmin must be greater than 0.
This phenomenon is expressed in Equation (13) as
√
1 M 4 2 1
1 1 1 − dZ
[ − ] I −√ VOUTrms d Z Ts > 0 (13)
2 1 − 2d Z 3π 1 − (4π f m )2 L Z CZ OUTrms 2 L Z ηM
From Figure 5a, the absolute value of load impedance expressed in Equation (14)
should be lower in value (but always positive) than the value calculated in Equation (14)
for CCM for the assigned parameters: dZ , LZ , and CZ , M = 1 − dZ .

η ML Z M 1 8 1

| ZLOAD | < (√ − ) (14)
(1 − d Z )d Z Ts 2 1 − 2d Z 3π 1 − (4π f m ) L Z CZ
2
(a)
(b)
Figure 5. (a) Maximum load impedance, and (b) minimum output current, that keeps the impedance
network in the continuous current mode.
47
Energies 2021, 14, 7272
As shown in Figure 5b, the minimum output current for CCM is given Equation (15) as
1 1− d Z
ML Z 1−2d Z VDC d Z Ts
IOUTrms > √
(15)
8 2
1−2d Z − M3π 1−(4π f m )2 L Z CZ
1 1
The impedance network (Figure 5b) operates in the CCM for the ZSI load current
IOUTrms higher than the value calculated from Equation (15) for assigned LZ = 1 mH and
three parameters: VDC , dZ , and CZ . The modulation index M has the assigned maximum
possible value M = 1 − dZ .
In Figure 6, the continuous current mode is illustrated where the output voltage of the
ZSI is undistorted.
(a)
(b)
Figure 6. CCM waveforms of (a) the ILZ coil current, ZSI output voltage, and inverter PWM pulses,
and (b) the undistorted inverter output voltage.
Figure 7 presents the DCM where two cases can be distinguished. From this figure, the
distortions of the output voltage are small when the output voltage is below the maximum.
When the output voltage is closer to the maximum, the distortions are higher, and the
output voltage maximum is lower than expected. For the large VSI output capacitor the VSI
output and PWM envelope voltages are shifted when the large VSI output capacitor e.g.,
CF = 50 μF is used. As shown in Figure 7, the short PWM pulses are undistorted in DCM
48
Energies 2021, 14, 7272
while the wide pulses are distorted, and the output voltage is lower. The simulation of a
DCM operation using the Z-source is presented in Figure 8 for the third PWM modulation
schema [30]. The variables used to obtain the measured plots in Figure 8 are given as:
CF = 1 μF, dZ = 0.3, M = 0.65, RLOAD = 1000 Ω, 3rd modulation schema.
Figure 7. Measured DCM waveforms of the LZ coil current, ZSI output voltage, and the inverter’s
PWM wide and short pulses for CF = 1 μF and 50 μF inverter capacitors.
Figure 8. Simulated DCM waveforms for inverter CF = 1 μF, dZ = 0.3, M = 0.65, RLOAD = 1000 Ω, 3rd
modulation schema.
3. Controlled Energy Flow—Charging the Battery

Similar results of measurement shown in Figure 7 and simulations in Figure 8 demon-
strate that further simulations of the controlled energy flow i.e., charging the battery is
useful. The basic solution is an efficient multi-input-single-output (MISO) [31] feedback
that can decrease total harmonic distortions (THD) [23,27]. In addition, MISO feedback can
decrease two other types of output voltage distortions [27]. However, for systems supplied
by varying the DC supply voltage, for example, photovoltaic cells, the controlled energy
flow to the batteries, which keeps the CCM, can be used. It is recommended that the battery
is charged with a current that is a function of the difference between the calculated value
of IOUTrmsmin and averaged (10 Hz low pass filter) VSI output current IOUTrms as shown in
49
Energies 2021, 14, 7272
Figure 9 (if this difference is negative the charging battery current is equal to zero). The
actual difference of these currents IOUTrmsmin − IOUTrms is recalculated (if positive) to match
the required increase of the average ILZav current expressed in Equation (7). The battery can
be charged only during the non-shoot-through state. Energy from the battery is discharged
when VDC decreases below the assumed value of VDCmin , the Z-source is switched off and
the shoot-through pulses are blocked.
Figure 9. Proposed idea of the inverter with the impedance network charging the battery from
the DC link (during non-shoot-through time), and automatic switching to supplying directly from
the battery (the positions of switches are presented in the position of discharging the battery when
VDCmin − VDC > 0).
The idea of this system is presented in Figure 9 (for switches placed in the position
of discharging the battery). When the battery returns energy, the following happens: the
shoot-through pulses are stopped, and the 48 V battery is connected directly to the VSI.
This battery voltage should be higher than the amplitude of the output sinusoidal voltage
and the modulation index M of VSI is increased i.e., M2 is greater than M1 (Figure 9).
Figure 10a presents the simulated waveforms of the VDC changed 24/12/24 V (the bor-
der value is set to 15 V) with the described automatic action from Figure 9 but without con-
trolled charging the battery when Z-Source operates in the DCM. The following parameters
were used in this scenario: dZ = 0.3, M1 = 0.65, M2 = 0.75 and RLOAD = 1000 Ω. Figure 10b
presents that same operation but with controlled charging of the battery for keeping Z-
Source in the CCM. The current charging of the battery is calculated as
IBATT = f (IOUTrmsmin − IOUTrms ) using Equation (15), where f is a function of Equation
(7). The battery charging current IBATT calculated from Equations (7) and (15) should
be reduced because too high a value of the battery charging current leads to distortions
of the VSI output voltage time after the output voltage is zero-crossing (see Figure 11b).
These distortions are caused by the high voltage drops on the VSI switches during the
shoot-through time. The presented (Figure 10b) reduction of the output voltage THD from
4.6% to 3% without any feedback loop is quite promising.
50
Energies 2021, 14, 7272
(a)
(b)
Figure 10. The waveforms of the DC input and AC output voltages of the ZSI switched from a mode
of supplying the VSI from Z-source to the mode of supplying VSI from the battery in case of the low
input DC voltage, (a) without controlled charging battery for Z-source in the DCM for the low load,
and (b) with controlled charging battery for Z-source in CCM.
The presented simulations were verified in an experimental model using a 12 V

battery (without discharging the battery) charged from the DC during dB Ts pulses where
(dB = 1 − dZ ) (Figure 12). The feedback loop was the IPBC2 type presented in [27]. For the
DCM mode of the Z-source, the output voltage distortions can be reduced by additional
loading the impedance network by means of charging the battery from the DC link in the
non-shoot through times.
51
Energies 2021, 14, 7272
Figure 11. Inverter output voltage (a) without charging battery, (b) the battery charging current
directly equal to f (IOUTrmsmin − IOUTrms ), where f is a Equation (7), and (c) the battery charging with
the reduced value of current.
(a)
Figure 12. Cont.
52
Energies 2021, 14, 7272
(b)
Figure 12. (a) The inverter experimental set up and (b) inverter output voltage distortions com-
parison for an IPBC controller where RLOAD = 2000 Ω, RMS battery charging currents: IBATT = 0
(DCM of the Z-source), IBATT = 120 mA and IBATT = 200 mA (CCM of the Z-source), dZ = 0.3, and
dB = 1 − dZ —battery charging pulses coefficient.
The current source from Figure 9 was simply substituted with resistors. Charging
the battery allowed for a substantial reduction of output voltage THD from 2.63% to 0.9%.
for IBATT = 120 mA, but THD increased to 0.97% for IBATT = 200 mA. Further research will
be on the use of battery charging current not only to reduce the distortions of the output
voltage but also looking for a maximum power point (MPP) when the impedance network
is supplied from the photovoltaic cell. The battery charging current can be controlled by
the coefficient dB for the input current of the impedance network would be closer to MPP.
4. Discussion
The presented results of the simulation and measurements of the experimental ZSI
proved that charging the battery from the DC link between impedance network and VSI
in the non-shoot-through time can seriously decrease the ZSI output voltage distortions
keeping the impedance network in the CCM. The controlled energy flow solution is
particularly predicted for the case of wide variations of the input DC voltage and variations
of the load current. The output voltage distortions are decreased even when a strong
feedback loop of the VSI is present. The controlled charging of the battery can help in the
maximum power point tracking when the ZSI is supplied from the photovoltaic cell and
this is the perspective of the further studies. In [23], three types of VSI output voltage
distortions were distinguished. The controlled charging of the battery can cancel one of
them but setting too high a value of this current increases the other reason for distortions.
Charging the battery from the DC link of the ZSI during the non-shoot-through time was
not presented yet, however, another approach to the controlled power flow for qZSI with
charging the battery connected parallel to the CZ2 capacitor (Figure 2) was presented in [32].
5. Conclusions
In this paper, a technique has been proposed to reduce output voltage distortions in
voltage source inverters connected to impedance networks. The proposed method has
been validated using simulations and experimentally under different operating conditions.
It was discovered that by connecting a rechargeable battery to a DC link placed between
53
Energies 2021, 14, 7272
an impedance network and a VSI and employing proper control of the battery charging
current during the non-shoot through time, the output voltage distortions in a system with
or without feedback can be reduced when a continuous current mode of the impedance
network is forced. However, too high a current charging the battery may increase other
types of VSI output voltage distortions presented in Figure 11b caused by high voltage
drops on the VSI switches during the shoot-through time. Furthermore, the battery charg-
ing current can be controlled to increase the impedance network input current to enable
the system to reach the maximum power point when the DC source is a photovoltaic cell.
The results presented in this paper thus demonstrate that the proposed method is suitable
and can be applied in practice to real-time supply systems.
Author Contributions: Conceptualization, Z.R.; methodology, Z.R. and K.B.; software, Z.R.; valida-
tion, Z.R., K.B. and Ł.D.; formal analysis, Z.R. and K.B.; investigation, Z.R. and K.B.; resources, Z.R.
and K.B.; data curation, Z.R. and K.B.; writing—original draft preparation, Z.R.; writing—review
and editing, Z.R. and K.B.; visualization, Z.R.; supervision, Z.R.; project administration, Z.R. and
K.B.; funding acquisition, Z.R. and K.B. All authors have read and agreed to the published version of
the manuscript.
Funding: This work was partly supported by the Polish Ministry of Science and Higher Education
funding for statutory activities.
Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design
of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or
in the decision to publish the results.
References
1. Peng, F.Z. Z-Source Inverter. IEEE Trans. Ind. Appl. 2003, 39, 504–510. [CrossRef]
2. Anderson, J.; Peng, F. Four quasi-Z-Source inverters. Proc. IEEE Power Electron. Spec. Conf. 2008, 2743–2749. [CrossRef]
3. Li, Y.; Peng, F.Z. AC Small Signal Modeling, Analysis and Control of Quasi-Z-Source Converter. In Proceedings of the 2012 IEEE
7th International Power Electronics and Motion Control Conference—ECCE Asia, Harbin, China, 2–5 June 2012; pp. 1848–1854.
[CrossRef]
4. Liu, W.; Yang, Y.; Kerekes, T.; Liivik, E.; Blaabjerg, F. Impedance Network Impact on the Controller Design of the QZSI for
PV Applications. In Proceedings of the 2020 IEEE 21st Workshop on Control and Modeling for Power Electronics (COMPEL),
Aalborg, Denmark, 9–12 November 2020. [CrossRef]
5. Subhani, N.; Kannan, R.; Mahmud, A.; Blaabjerg, F. Z-source inverter topologies with switched Z-impedance networks: A review.
IET Power Electron. 2021, 14, 727–750. [CrossRef]
6. Zhu, M.; Yu, K.; Luo, F.L. Switched Inductor Z-Source Inverter. IEEE Trans. Power Electron. 2010, 25, 2150–2158. [CrossRef]
7. Itozakura, H.; Koizumi, H. Embedded Z-Source Inverter with Switched Inductor. In Proceedings of the IECON 2011—37th
Annual Conference of the IEEE Industrial Electronics Society, Melbourne, VIC, Australia, 7–10 November 2011; pp. 1342–1347.
[CrossRef]
8. Adamowicz, M.; Guzinski, J.; Strzelecki, R.; Peng, F.Z.; Abu-Rub, H. High Step-Up Continuous Input Current LCCT-Z-Source
Inverters for Fuel Cells. In Proceedings of the Energy Conversion Congress and Exposition (ECCE), Phoenix, AZ, USA, 17–22
September 2011; pp. 2276–2282. [CrossRef]
9. Adamowicz, M. LCCT-Z-Source Inverters. In Proceedings of the 10th International Conference on Environment and Electrical
Engineering, Rome, Italy, 8–11 May 2011; pp. 1–6. [CrossRef]
10. Vinnikov, D.; Roasto, I.; Strzelecki, R.; Adamowicz, M. Step-Up DC/DC Converters With Cascaded Quasi-Z-Source Network.
IEEE Trans. Ind. Electron. 2012, 59, 3727–3736. [CrossRef]
11. Siwakoti, Y.P.; Blaabjerg, F.; Galigekere, V.P.; Ayachit, A.; Kazimierczuk, M.K. A-Source Impedance Network. IEEE Trans. Power
12. Hakemi, A.; Sanatkar-Chayjani, M.; Monfared, M. Δ-Source Impedance Network. IEEE Trans. Ind. Electron. 2017, 64, 1–10.
[CrossRef]
13. Rezazadeh, H.; Monfared, M.; Nikbahar, A.; Sharifi, S. A family of high voltage gain quasi-Δ-source impedance networks. IET
14. Siwakoti, Y.P.; Peng, F.Z.; Blaabjerg, F.; Loh, P.C.; Town, G.E. Impedance-Source Networks for Electric Power Conversion Part I: A
Topological Review. IEEE Trans. Power Electron. 2015, 30, 699–716. [CrossRef]
15. Reddivari, R.; Jena, D. A Correlative Investigation of Impedance Source Networks: A Comprehensive Review. IETE Tech. Rev.
2021, 38, 1–34. [CrossRef]
16. Ghasimi, S.; Eshkevari, A.L.; Mosallanejad, A. A high-gain IΓ-source hybrid single-phase multilevel inverter for photovoltaic
application. IET Power Electron. 2021, 14, 106–119. [CrossRef]
54
Energies 2021, 14, 7272
17. Kumar, A.; Bao, D.; Beig, A.R. Comparative Analysis of Extended SC-qSBI with EB-QZSI and EB/ASN-QZSI. IEEE Access 2021, 9,
61539–61547. [CrossRef]
18. Zhao, P.; Wang, J.; Hao, H.; Wang, U. Y-Source Two-Stage Matrix Converter and Its Modulation Strategy. IEEE Access 2020, 8,
214282–214292. [CrossRef]
19. Yuan, J.; Mostaan, A.; Yang, Y.; Siwakoti, Y.P.; Blaabjerg, F. A Modified Y-Source DC–DC Converter With High Voltage-Gains and
Low Switch Stresses. IEEE Trans. Power Electron. 2020, 35, 7716–7720. [CrossRef]
20. Gayen, P.K. An enhanced high-boost active-switched quasi Z-Source inverter having shorter range of shoot-through duty ratio
for solar energy conversion applications. AEU—Int. J. Electron. Commun. 2021, 137, 153822. [CrossRef]
21. Siwakoti, Y.P.; Peng, F.Z.; Blaabjerg, F.; Loh, P.C.; Town, G.E.; Yang, S. Impedance-Source Networks for Electric Power Conversion
Part II: Review of Control and Modulation Techniques. IEEE Trans. Power Electron. 2015, 30, 1887–1906. [CrossRef]
22. Abdelhakim, A.; Blaabjerg, F.; Mattavelli, P. Modulation Schemes of the Three-Phase Impedance Source Inverters—Part I:
Classification and Review. IEEE Trans. Ind. Electron. 2018, 65, 6309–6320. [CrossRef]
23. Rymarski, Z.; Bernacki, K. Drawbacks of impedance networks. Int. J. Circuit Theory Appl. 2018, 46, 612–628. [CrossRef]
24. Bernacki, K.; Rymarski, Z. Electromagnetic Compatibility of Impedance Source Inverters. Elektron. Elektrotechnika 2017, 23, 55–63.
[CrossRef]
25. Hufman, B. Efficiency and Power Characteristics of Switching Regulator Circuits. Linear Technol. Appl. Note 1991, 46, 1–28.
26. Ge, B.; Abu-Rub, H.; Peng, F.Z.; Lei, Q.; Almeida, A.T.; Ferreira, F.J.T.E.; Sun, D.; Liu, Y. An Energy-Stored Quasi-Z-Source Inverter
for Application to Photovoltaic Power System. IEEE Trans. Ind. Electron. 2013, 60, 4468–4481. [CrossRef]
27. Rymarski, Z.; Bernacki, K.; Dyga, Ł. Decreasing the single phase inverter output voltage distortions caused by impedance
networks. IEEE Trans. Ind. Appl. 2019, 55, 7586–7594. [CrossRef]
28. Rymarski, Z.; Bernacki, K. Influence of Z-Source output impedance on dynamic properties of single-phase voltage source inverters
for uninterrupted power supply. IET Power Electron. 2014, 7, 1978–1988. [CrossRef]
29. Bernacki, K.; Rymarski, Z.; Dyga, Ł. Selecting the coil core powder material for the output filter of a voltage source inverter.
Electron. Lett. 2017, 53, 1068–1069. [CrossRef]
30. Bernacki, K.; Rymarski, Z. Electromagnetic compatibility of voltage source inverters for uninterruptible power supply system
depending on the pulse-width modulation scheme. IET Power Electron. 2015, 8, 1026–1034. [CrossRef]
31. Astrom, K.J.; Wittenmark, B. Computer-Controlled Systems: Theory and Design, 3rd ed.; Dover Publications Inc.: Mineola, NY, USA,
2011; ISBN 9780486486130.
32. Sun, D.; Ge, B.; Peng, F.Z.; Abu Rub, H.; de Almeida, A.T. Power flow control for quasi-Z source inverter with battery based PV
power generation system. In Proceedings of the 2011 IEEE Energy Conversion Congress and Exposition, Phoenix, AZ, USA,
17–22 September 2011; pp. 1051–1056. [CrossRef]
55
energies
Review
A Comprehensive Review of Lithium-Ion Cell Temperature
Estimation Techniques Applicable to Health-Conscious Fast
Charging and Smart Battery Management Systems
Akash Samanta 1, * and Sheldon S. Williamson 2
1 Department of Applied Physics, Faculty of Electrical Engineering, University of Calcutta,

Kolkata 700009, India
2 Department of Electrical, Computer and Software Engineering, Faculty of Engineering and Applied Science,
Ontario Tech University, Oshawa, ON L1G 0C5, Canada; Sheldon.Williamson@ontariotechu.ca
* Correspondence: akashsamanta440@gmail.com; Tel.: +91-9143877405
Abstract: Highly nonlinear characteristics of lithium-ion batteries (LIBs) are significantly influenced
by the external and internal temperature of the LIB cell. Moreover, a cell temperature beyond the
manufacturer’s specified safe operating limit could lead to thermal runaway and even fire hazards
and safety concerns to operating personnel. Therefore, accurate information of cell internal and
surface temperature of LIB is highly crucial for effective thermal management and proper operation
of a battery management system (BMS). Accurate temperature information is also essential to BMS for
the accurate estimation of various important states of LIB, such as state of charge, state of health and so
on. High-capacity LIB packs, used in electric vehicles and grid-tied stationary energy storage system
essentially consist of thousands of individual LIB cells. Therefore, installing a physical sensor at each
Citation: Samanta, A.; Williamson,
cell, especially at the cell core, is not practically feasible from the solution cost, space and weight
S.S. A Comprehensive Review of
point of view. A solution is to develop a suitable estimation strategy which led scholars to propose
Lithium-Ion Cell Temperature
Estimation Techniques Applicable to
different temperature estimation schemes aiming to establish a balance among accuracy, adaptability,
Health-Conscious Fast Charging and modelling complexity and computational cost. This article presented an exhaustive review of these
Smart Battery Management Systems. estimation strategies covering recent developments, current issues, major challenges, and future
Energies 2021, 14, 5960. https:// research recommendations. The prime intention is to provide a detailed guideline to researchers
doi.org/10.3390/en14185960 and industries towards developing a highly accurate, intelligent, adaptive, easy-to-implement and
computationally efficient online temperature estimation strategy applicable to health-conscious fast
Academic Editors: Andrei Blinov, charging and smart onboard BMS.
Sheldon Williamson, Seung-Wan
Song and Mario Marchesoni Keywords: electric vehicles; machine learning; Kalman filter; thermal modelling; online prediction;
electromagnetic impedance spectroscopy; computational cost
Received: 5 July 2021
Accepted: 13 September 2021
Published: 20 September 2021
1. Introduction
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
Lithium-ion batteries (LIBs) are widely used in electric vehicles (EVs), grid-tied sta-
published maps and institutional affil- tionary energy storage systems, and several other consumer electronics primarily due
iations. to their high voltage rating (>4 V/cell) and high energy density (~265 (W h) L−1 ) and
longer operational life. The use of LIBs in automotive and aerospace applications has led to
larger cell sizes and large battery packs for a higher driving range and the requirement for
more aggressive charging and discharging. However, thermal instability and temperature-
dependent nonlinear behavior is some of the common concerns behind the safe and reliable
operation of LIB systems. It is noticed that the operation of batteries outside the safe
operating temperature directly affects the performance of LIBs, such as cycle life, efficiency,
distributed under the terms and reliability and safety. Researchers investigating the thermal performance of LIB showed
conditions of the Creative Commons that the best operating temperature range is from 25 ◦ C to 40 ◦ C [1,2]. Richardson et al. [3]
Attribution (CC BY) license (https:// demonstrated that the difference between the core and surface temperature could reach
creativecommons.org/licenses/by/ more than 10 ◦ C during real-life applications, especially during the high discharging condi-
4.0/). tion and fluctuating load current demand. The excessive temperature difference and the

Energies 2021, 14, 5960
accumulation of a large amount of heat inside the cell could lead to thermal runaway or
even explosions and fire [4]. That necessitates the employment of a battery management
system (BMS) for effective monitoring of battery parameters (current, voltage, temperature),
estimation of battery states (state of charge (SOC), state of health (SOH), remaining useful
life (RUL), state of temperature (SOT) [5]). Research studies demonstrated that SOC [6],
SOH [7], and remaining storage capacity [8] are a function of temperature; thus, the esti-
mation of the battery states also depends on the accurate estimation of cell temperature.
The Columbic efficiency of a cell is greatly affected by the cell temperature during the
charging and discharging period. Few other popular functionalities of BMS include cell
balancing [9] and fault detection/diagnosis [10] to ensure optimum capacity utilization,
operational safety, reliability, and longer battery life often requires temperature information
of an individual cell and battery pack as well. Therefore, accurate information of core and
surface temperature is highly crucial for effective thermal management and safety of a
LIB pack. Moreover, in cold climate areas, the battery capacity is drastically reduced due
to low-temperature operation that requires preheating the battery to a suitable range for
optimum performance [11,12]. It is also evidenced that for every 0.1 ◦ C beyond the safe
operating region the battery capacity degrades by about 5% [13]. It is evidenced that maxi-
mum heat is generated during the discharging period especially with fast discharging [14].
Therefore, accurate temperature estimation is essential for effective thermal management
and safety during fast charging and discharging and preheating of the cell to minimize
capacity fade.
In summary, it could be stated that the accurate information of cell temperature is
undoubtedly serving as the essential basis for the thermal management and safety of LIB.
While the surface temperature of each cell can be measured by installing a temperature
sensor on each cell, the core or internal temperature measurement directly using physical
sensors is challenging. Moreover, installing a temperature sensor on each cell surface is not
practically feasible from a system cost, space and weight point of view as any high-capacity
battery pack used in EVs and grid-tied systems essentially consists of thousands of individ-
ual cells. Researchers have also incorporated multi-dimensional sensing and self-healing
functions into a single battery cell to develop a smart battery [15–18]. Smart cells are typi-
cally capable of parameter measurements and estimation of cell states including the state
of temperature. Despite the modularized application of BMS in smart batteries, accurate
temperature estimation is still required, as otherwise installing sensors in each cell results
in high implementation cost and complexity. Therefore, researchers are struggling hard
to develop a high-fidelity, accurate, easy-to-implement, and computationally inexpensive
online temperature estimation strategy suitable for low-cost onboard BMS. Several tem-
perature estimation techniques have been proposed by researchers so far. Each different
type of method has its advantages and limitations with respect to the above-mentioned
features of an optimum BMS. Therefore, a summary of all the prominent techniques would
be very helpful to researchers and developers serving as a baseline for further research
and as a guideline for selecting appropriate techniques suitable for a specific requirement.
However, such a summary with detailed discussion on current progress and explanation of
the existing issues, challenges and future research scopes has not yet been presented in the
literature. Therefore, this article covered the research gap by conducting a comprehensive
review of the state-of-the-art temperature estimation strategies reported in the literature
so far.
The paper is organized as follows: In Section 2, generic temperature estimation strat-
egy of LIB is presented. The classification of temperature estimation strategies is presented
in Section 3. Section 4 is dedicated to presenting the existing estimation techniques, their
evolutions, limitations and challenges. It should be noted that temperature estimation
strategies for LIBs reported in the literature between 2010 to 2021 are primarily consid-
ered. However, few prominent research articles published between 1990 to 2010 are also
considered for understanding the fundamentals and evolution of temperature estimation
schemes. Commonly used search platforms, such as “Google Scholar”, “Science Direct”,
58
Energies 2021, 14, 5960
and IEEE Xplore, were used to find research articles published within this tenure. The
search criteria were “Temperature Estimation of Lithium-ion Batteries”. Section 5 discusses
the current issues, challenges and future research recommendations. Finally, Section 6 is
dedicated to a summary of the major findings and concluding remarks.
2. Generic Temperature Estimation Strategy

Irrespective of battery chemistry, heat is accumulated inside the battery during the
charging/discharging even during idle conditions, majorly due to several largely exother-
mic chemical and electrochemical reactions as well as transport processes. If the heat trans-
fer from the battery to the surroundings is not sufficient, then the heat gets accumulated
inside the battery resulting in an increase in core and surface temperature, thereby risking
thermal runaway. This phenomenon is even more prominent in the case of hard-cased
insulated batteries (as used in EVs), under fast charging/discharging and the operation
in hot environments. Heat dissipation is worse in cylindrical LIBs that are extensively
used in high-capacity LIB packs. Therefore, a typical temperature estimation scheme
consists of two models, namely, a heat generation model and a heat transfer model [19].
Often, a battery electrical model is also used to estimate the total heat generation using
Bernardi’s [20] heat generation model whereas few other models use a mathematical form
of battery electrochemistry to calculate the heat generation. Adaptive estimation strategies
also consider the influence of different battery states, such as SOC and SOH, as the battery
temperature is a function of these battery states. Then, the heat transfer model takes the
estimated total heat quantity as well as few other external measurements such as ambient
temperature to predict the temperature of that cell. Closed-loop estimation schemes use the
measured or the estimation temperature as feedback to improve the prediction accuracy. A
schematic layout of a generic temperature estimation strategy for LIB is shown in Figure 1.
Figure 1. Schematic layout of a generic temperature estimation strategy for a LIB cell.
3. Classification of Temperature Estimation Strategies

As shown in Figure 1, typically, a temperature estimation scheme consists of a heat
generation model and a heat transfer model. The heat generation models reported in the
literature can be broadly classified from two different aspects; based on modelling strategy
and based on the source of heat generation. Heat generation models based on modelling
strategy can be classified into three groups, physics-based electrochemical models [21–24],
equivalent circuit models (ECM) [25–27], black-box models [28–30]. In contrast, based
on the source of heat generation, these models can be grouped as a concentrated model,
distributed model [31] and heterogeneous model [25,32]. The concentrated heat generation
model considers that all heat is generated at the core only, usually considered to reduce the
modelling complexity. The distributed heat generation model considers that uniform heat
is generated throughout the entire cell geometry whereas the heterogeneous model can
capture different heat generation from difference cell layers usually resulting in temperature
and current density gradients inside the cell. The heterogeneous models are more detailed
thus can produce highly accurate predictions; however, these are most complex and require
extensive experiments for modelling. Distributed heat generation models are a balance
59
Energies 2021, 14, 5960
between the concentrated and heterogeneous models. The heat transfer models can be
classified into finite element analysis (FEA)-based models [27,33–36], heat capacitor-resistor
models (lumped or distributed parameter) [28,37–40], and data-driven techniques. Heat
capacitor–resistor-based models use the analogy between electrical and thermal systems.
A heat capacitor–resistor can be further classified as mentioned in Figure 2. Lumped
parameter models are simple and useful for online applications, however, only one or two
average temperatures can be predicted with these models whilst the battery temperature
distribution is not spatially uniform, especially in larger capacity cylindrical LIB cells. On
the other hand, complex distributed models [41,42] can describe the detailed temperature
distribution in a cell, however, they are not suitable for online application due to their
computational complexity. Several other detailed models of LIB accounting for the thermal
characteristics of different layers are studied in [43–48]. A two-state/node model provides
information on core and surface temperature whereas a one-state/node model can provide
only core temperature.
Figure 2. Family of (A) Heat generation model, (B) Heat transfer model, (C) Temperature estimation strategy.
The heat transfer model where the total heat generation is one of the input parameters
is collectively called the battery thermal model where the total heat generation is estimated
by the battery heat generation model. The thermal modelling of LIB is a separate area
of study and is not under the scope of this study. It deals only with the temperature
estimation strategies. However, as most of the temperature estimation strategies are
extensively depending on thermal modelling, an overview of each modelling technique is
60
Energies 2021, 14, 5960
also discussed with the respective temperature estimation strategy for better understanding.
Researchers employed different types of heat generation models with different kinds of
heat transfer models to come up with a temperature estimation scheme. Therefore, it is
challenging to classify these estimation strategies. Broadly, the temperature estimation
schemes can be grouped into electrochemical thermal modelling-based, equivalent electric
circuit model (EECM)-based, machine learning (ML)-based, numerical-model based, direct
impedance measurement-based, magnetic nanoparticles-based schemes. The families of
the LIB heat generation model, heat transfer model and temperature estimation strategy
are illustrated in Figure 2.
4. Comprehensive Review of Temperature Estimation Strategies

4.1. Electrochemical Thermal Modelling-Based Temperature Estimation
Researchers started thermal modelling in the early nineties, those are mostly coupled
with an electrochemical model to simulate the temperature profile of a battery under differ-
ent operating conditions, geometries or cooling rates. There are simple one-dimensional
(radial direction) models [37,49–54] to complex three-dimensional thermal models [55–59].
Researchers have primarily used different analytical techniques to mathematically model
the electrochemical behavior of the cell. One-dimensional models typically assume isother-
mal, constant current operation of the battery and lumped thermophysical properties
and constant heat generation rates. Highly complex three-dimensional models require an
in-depth understanding of the thermodynamic properties of battery materials and parts to
consider the heat effects caused by ohmic resistance, chemical reactions, mixing processes,
polarization and electrode kinetic resistance. Often, temperature estimation using such
highly complex models is very accurate, however, such detailed models are essential for
battery design purposes. Those are not compatible with temperature estimation using
onboard BMS with low computational resources. These complex models are capable of
accounting for the time-varying nonlinear battery performance. However, they typically
require several system properties, operational parameters which require extensive exper-
imental measurements. While, at the same time, quantitative estimation of some of the
properties, such as transport properties, thermodynamic properties and heat effects are
highly challenging.
Thomas and Newman [60] introduced an electrochemical modelling-based detailed
heat generation model of LIB to estimate the total heat generation during the charg-
ing/discharging period. The fundamental equation of the total heat generation inside the
LIB cell as proposed by Thomas and Newman reads

∂U avg ∂Cj
− ∑i ΔHi ri − ∑ j ( H j − H avg
avg
Q = I (V − U avg ) + IT j ) dv (1)
∂T ∂t
In Equation (1), Q is the rate of heat generated or consumed inside the cell, V and U
are the cell voltage and equilibrium potential, respectively, I is the charging or discharging
current, T is the cell temperature. ΔHi represents the changes in enthalpy of the chemical
reaction i and ri is the rate of reaction i. ΔH j represents the partial molar enthalpy of
species j and c j is the concentration of the species. t and v represent the time and volume
of the cell, respectively. All the properties are mentioned based on the volume-averaged
concentration, thus the superscript “avg” is used. The model can provide accurate informa-
tion on heat generation only, temperature estimation is not presented in this study. Their
heat generation model was extensively used by several other researchers. Modelling is
very detailed thus highly complicated and not suitable for online application owing to
the computational burden. One of the widely-used electrochemical models commonly
known as the Doyle–Fuller–Newmann model [37,61] is extensively referred to and also
used for thermal modelling. It consists of nonlinear partial differential-algebraic equations
to describe the internal characteristics of LIB. It is also referred to as the pseudo-two-
dimensional (P2D) model. The major limitation of the model is its high computational
burden which limits its application in online state estimation in embedded BMSs. Here,
61
Energies 2021, 14, 5960
Al Hallaj et al. [53] showed that a simplified transient one-dimensional thermal model
with lumped parameters is sufficient for cell design purposes, especially to simulate the
thermal behavior of scaled-up LIBs. Detailed knowledge of the role of different cell com-
ponents, such as electrodes, electrolytes and separators in heat generation is also not
necessary. Few researchers used this type of complex electrochemical model to explore
pulse power limitations to prevent thermal runaway and to design thermal management
systems [62,63]. Those are mostly used for designing LIB cells as well as LIB packs. A
lumped electrochemical-thermal-coupled model was used to predict the thermal perfor-
mance of LIB alongside the performance of individual electrodes at various operating
temperatures by Fang et al. [64]. The model was validated against the experimental data
for constant current and pulsing conditions characteristic of hybrid electric vehicle (HEV)
which are merely providing the laboratory experimental results instead of a real-work
application scenario. The impact of charging current on internal temperature behavior
was investigated in [65]. Gerver et al. [66] included more detailed information and cell
characteristics to develop a multi-dimensional electrochemical thermal model of LIB to
analyze the thermal performance and heat generation more accurately. Despite estimation
accuracy, the modelling complexity and computational burden limit its application in
embedded BMS.
Due to a lack of clear understanding of the electrochemical processes inside the LIB
and their corresponding mathematical equations alongside to reduce the computational ex-
penses, often all heat generation sources were not modelled/considered. These unmodeled
heat generation behaviors lead to significant errors in temperature estimation. Regarding
this, Zhang et al. [67] developed a two-state thermal model utilizing discretization and
inverse model techniques which do not require prior knowledge of thermal boundary
conditions. Moreover, the model is capable of estimating the total heat generation of a
battery cell, thus, thermal modelling of each heat source is not required and abnormal
heat generation can also be detected from the estimation results. The effectiveness and
robustness of the model were tested for varying thermal boundary conditions and fast
charging conditions. While the strategy is designed for self-heating pouch cells, a sim-
ilar approach could also be adapted for other types of LIBs. Thus, further research is
recommended here. A high-fidelity electrochemical model and onboard measurements
such as terminal voltage and current were used by Wang et al. [68] to estimate the cell
temperature at a wide range of C-rates during the charging/discharging period. They
have also used a dual ensemble Kalman filter (DEKF) which incorporates enhanced single-
particle dynamics to relate terminal voltage to battery temperature and Li+ concentration.
Besides, modelling complexity and high computational cost, the accurate determination
of lithium (Li+) concentration is challenging. Therefore, the application of the model in
real-life online prediction is questionable. The spatial distribution of internal temperature
in LIB was estimated using a pseudo-2D electrochemical model and soft-constrained dual
unscented Kalman filter (DUKF) by Marelli and Corno [69]. It is mainly developed to
estimate the Li+ concentration and modelling complexity and computational expenses
are very high. However, the approach could be extended for temperature estimation.
Smith et al. [62] developed a one-dimensional electrochemical, lumped thermal model to
explore pulse power limitations and thermal behavior of a LIB pack. The electrochemical
thermal modelling-based temperature estimation strategies proposed by different authors
are summarized in Table 1 for a quick reference to the readers. In general, the major
limitations of any electrochemical model-based strategies are the modelling complexity
and high computational cost making these models unsuitable for online prediction and
application at low-cost onboard BMS.
62
Energies 2021, 14, 5960
Table 1. Summary of electrochemical thermal modelling-based temperature estimation strategies.
Reference Types of Models Important Note

One-dimensional
Thomas and Newman [60] Not used for temperature estimation
electrochemical model
Doyle–Fuller–Newmann model Pseudo-two-dimensional Not used for temperature estimation but several other
[37,61] (P2D) model researchers used
A transient one-dimensional
Detailed information of electrodes, electrolytes and
Al Hallaj et al. [53] thermal model with
separator were considered in heat generation model
lumped parameters
Can estimate one or two average temperatures,
Lumped parameter
performance of individual electrode at various operating
Fang et al. [64]. electrochemical-thermal-
temperatures, constant current and pulsing conditions
coupled model
characteristic were considered, experimentally validated
Temperature-dependent physicochemical properties and
Thermal energy generation model, thermal behaviors under various charging conditions
Gu and Wang [41] multiphase micro-macroscopic were considered. Capable of predicting the average cell
electrochemical model temperature as well as the temperature distribution inside
a cell, volume-averaging technique, numerical simulations
Thermal dependence of various parameters in the model
Kumaresan et al. [42] One-dimensional thermal model on different discharge profiles was assessed, validated
using experimental and simulation results
Able to provide temperature distribution based on
Two-dimensional modelling +
Kim et al. [65] potential and current density distribution, MATLAB,
Finite element method (FEM)
validated using experimental and simulation results
A multi-dimensional Thermal properties of each cell layer are considered,
Gerver et al. [66]
electrochemical thermal model experimentally validated
High-fidelity electrochemical
Wide range of C-rates during the charging/discharging
model + onboard measurements +
Wang et al. [68] period, MATLAB, validated using experimental and
dual ensemble Kalman
simulation results
filter (DEKF)
Pseudo-2D electrochemical model
Can provide information on the spatial distribution of
Marelli and Corno [69] and soft-constrained dual
internal temperature, MATLAB Simulation
unscented Kalman filter (DUKF)
A one-dimensional Adaptive to different drive-cycles, tested and validated
Smith et al. [62] electrochemical lumped with FUDS and HWFET drive cycles,
thermal model experimentally validated
4.2. Equivalent Electric Circuit Model-Based Temperature Estimation

An equivalent electric circuit model (EECM) represents the thermal dynamics of
LIB using electrical system parameters to develop a heat capacitor–resistor-based battery
thermal model. Depending on the number of heat capacitors (number of energy storage
elements) two types of models, namely, the first-order model and second-order model have
been developed so far in the literature. The first-order model consists of one thermal energy
storage element whereas a second-order thermal model consists of two heat capacitors,
typically, one for the heat capacitance of the core and the other one is for the cell surface [13].
The second-order model can capture more dynamics than the first-order model. The first-
order and second-order thermal models of a LIB cell are shown in Figure 3a,b, respectively.
In Figure 3, Q represents the heat generation rate, Cc and Cs are the heat capacitance of core
and surface, respectively, Tin and Tout are the temperatures of core and surface of the cell,
respectively. Tamb is the ambient temperature.
63
Energies 2021, 14, 5960
(a) (b)
Figure 3. Thermal model of a LIB cell (a) First-order model, (b) Second-order model.
Further, depending on the modelling complexity, EECM could be also classified as

lumped-parameter and distributed parameter models. Lumped-parameter models are
used for simplification and thus low computational cost compared to detailed distributed
models. Computationally efficient lumped thermal models are developed using single
temperature as input to capture the model parameters [70] while some researchers used
both surface and core temperatures of the cell to construct the lumped thermal models.
Some also considered the correlation between cell geometry and other physical properties
with thermal modelling [71]. However, several assumptions were made during modelling
leading to inaccurate temperature estimation compared to detailed thermal modelling.
Further, thermal models that only estimate the core temperature are considered as single-
state/node [72], whereas if the model can estimate both surface and core temperature
then it is termed as two-state/node [67] thermal model. The parameters of the EECM
are identified through ranges of experimental studies such as electrochemical impedance
spectroscopy (EIS) or utilizing externally measurable quantities, such as voltage, current,
and temperature. Few studies also considered various conditions of SOC, SOH and
estimated surface/core temperatures to make the model more robust. It is very difficult
to group those thermal models because lumped models are used in both single-state and
dual-state modelling and the model could be first-order and second-order. Therefore, the
literature is grouped into cell-level and pack-level temperature estimation schemes that are
discussed below.
Typically, these EECM models determine the value of Q using Equation (2) as formu-
lated by Bernardi et al. [20]:
dVocv
Q = I (V − VOCV ) + ITc (2)
dTc
where Vocv represents the open-circuit voltage of the battery cell and the term dV ocv
dTc is the
entropy coefficient. Finally, Tc and Ts are estimated using the mathematical form of the thermal
models shown in Figure 3. Mathematical equations for temperature estimation using the first-
order and second-order thermal model are represented by Equations (3) and (4), respectively.
4.2.1. EECM-Based Cell Temperature Estimation

One of the prime challenges of any EECM-based strategy is model parameter iden-
tification. Forgeze et al. [43] used transient experiments by applying current pulses of
different magnitudes to increase the internal temperature and the model parameters, heat
transfer coefficients and heat capacity were determined to construct a lumped parameter
thermal model. This study used EIS for parameter identification where current pulses at
2 Hz were used to increase the internal temperature. The Tc was estimated based on the
measured surface temperature using the lumped parameter thermal model. The entropy
change was also taken into account while modelling. They developed a first-order thermal
64
Energies 2021, 14, 5960
model as shown in Figure 3a. The mathematical representation of the first-order thermal
model as used by Forgeze et al. reads

R R
Tin = Ts 1 + in − Tamb in (3)
Rout Rout
The strategy developed by Forgez et al. lacks quantitative analysis of the influence
of heat generation. The operating current is much higher compared to the very low
current value used in EIS. Therefore, model parameters determined using EIS are not
appropriate for capturing the thermal dynamics accurately. Moreover, they have considered
uniform internal temperature, however, more than 10 ◦ C temperature difference among
different internal points of a cell has been reported in the same study. This strategy
requires surface temperature measurement by installing a temperature sensor at each cell,
thus scaling-up is impractical. Maleki and Shamsuri [73] developed a thermal model of
notebook computer LIB-pack to understand the thermal response under various operating
conditions aiming to reduce the battery pack designing cost and time. They revealed that
the temperature rise during charging is dominated by heat dissipation from the control
power electronics while during discharging it is dominated by the heat generated inside
the LIB cell. These relevant observations must be considered while designing an effective
thermal management system of LIB pack, especially for health-conscious fast charging.
Surya et al. [13] developed a second-order thermal model for core and surface temperature
estimation scheme using KF. Here, the least square (LS) algorithm was employed to identify
the battery thermal parameters. Despite the simplicity and good accuracy, environmental
uncertainties were not considered during modelling. Moreover, they presented the results
based on simulation study alongside very simple and low-current discharge profile was
used for model validation, thus, the accuracy in the real-world applications needs further
investigation. Previously, models were validated using a simple charging/discharging
current profile. However, the load profile in real-life applications much deviates from
those simple loading profiles. Therefore, a second-order thermal model and ECM-based
two-state thermal model of cylindrical LIB cell were validated with two basic drive-cycle
tests, covering an SOC range 25–100%, temperature 5–38 ◦ C, and maximum C-rate of 22 by
Lin et al. [74]. The influence of the constantly varying temperature and SOC on the EECM
parameters and consequential effect on battery thermal performance was investigated
by Lin et al. [74]. The model demonstrated good prediction accuracy and robustness.
However, testing using standard internationally referred drive-cycles was not conducted.
Thus, accuracy and robustness in practical scenarios need further investigation. EECM
parameters are influenced by cell ageing, thus, Li and Yang [75] considered the influences
of ageing and heat transfer conditions on thermophysical model parameters. Li and
Yang identified the parameters of the extended lumped parameter model online where a
forgetting factor recursive least squares (FFRLS) algorithm was employed.
Further to this research, the uncertainties in practical operation were considered by
Lin et al. [45,76] alongside the impact of cell ageing during online parameter identification.
As an up-gradation, the commonly deployed LS algorithm was augmented with non-
uniform forgetting factors to track the time-varying internal parameters making the model
adaptive to cell ageing and other uncertainties. In [77], only two lumped models were
used to approximate the core and surface temperatures, respectively, which may not be
suitable for a large capacity LIB pack due to strong spatiotemporal thermal distribution.
While the influence of overpotential entropy changes on battery heat generation was
considered, core temperatures estimation of only a single cell was considered. Sun et al. [78]
developed a second-order lumped parameter thermal model with the KF technique for
core temperature estimation only (single-state). They used an ECM-based heat generation
model to mathematically model the accumulation of the total heat generation at the cell
core. As an improvement of previous studies, this study considered the influence of
entropy changes and overpotential on cell thermal behavior and was quantitively analyzed
to develop an online internal temperature estimation strategy. This strategy utilized surface
65
Energies 2021, 14, 5960
and ambient temperature for core temperature estimation during charge and discharge
cycles where the KF was used for adaptive estimation by the process of state and time
update in real-time. The impact of unmeasurable modelling error, the initialization error
and the possible time-varying external thermal resistance on the temperature estimation
accuracy were considered by Dai et al. [79]. In that paper, a second-order lumped parameter
thermal model, as shown in Figure 3b, was developed for adaptive core temperature
estimation based on the KF. Further, joint Kalman filtering (JKF) was used to simultaneously
estimate both core temperature and time-varying external thermal resistance online. The
mathematical equation employed by Dai et al. for core temperature estimation can be
represented as Equation (4):

1 + RRout
in
+ Cs Rin s
Tin (s) = Q(s) (4)
Cs Cc s2 + Cs + Cc + RRinout
Cc
s + R1out
where s is the Laplace operator. Other parameters are the same as mentioned in Figure 3.
The LS algorithm based on the experimental data was also used to determine the
lumped parameters of the thermal model. Dai et al. enhanced the modelling accuracy by
constructing a separate thermal model for core and battery shell alongside considering
the external heat exchange coefficient as time-varying. The authors simply stated that
the proposed method computes efficiently, however, no information about computation
time, hardware requirement was presented. Several assumptions were also made during
modelling, leading to inaccurate estimation in real-life applications.
A trade-off between the detailed and lumped parameter thermal modelling ap-
proaches was considered by Doughty et al. [80] and Park et al. [71]. They developed
a two-state thermal model that predicts the surface and core temperature of LIB. The novel
intention was to provide more information compared to the lumped model while reducing
the computational cost. Few researchers also termed the lumped parameter model as
a reduced-order model (ROM). Whilst the primary intention is same, that is, to reduce
the complex thermal problem into a simplified heat transfer problem characterized by a
reduced set of thermal parameters. A combination of lumped parameter two-state thermal
model with 2RC (second-order) ECM along with a joint Kalman filter (JKF)-based core and
surface temperature estimation strategy was proposed by Chen et al. [72]. The simulation
and experimental test were conducted to verify the adaptiveness of the model to constantly
varying temperature and SOC and, finally, the prediction accuracy was also assessed. It
was also demonstrated that the proposed model has higher prediction accuracy compared
to previously discussed EECMs. It was also demonstrated that the model is highly robust
against automatic correction for surface thermal resistance.
To provide more detailed information on the temperature distribution in cylindrical
LIB, Xie et al. [81] developed a one-dimensional (radial) lumped parameter thermal model
with a dual Kalman filter (DKF). As an improvement, this model is capable to provide
temperature information at three different points of the battery, compared to only core
and surface temperature. Thus, the researchers termed this modelling as a three-node
thermal model. In this study, the anisotropy of thermal conductivity was also considered in
identifying internal resistance and SOC during the temperature estimation to enhance the
prediction accuracy and robustness. The impact of different charging/discharging current
conditions was not considered. Moreover, 1-RC ECM-based heat generation model is
considered, thus presumably, the accuracy can be further improved with the application of
the 2-RC ECM-based heat generation model. Online parameter estimation using a particle-
swarm algorithm with pulse discharge experiments under different ambient temperatures
was employed by Pan et al. [19]. A combination of 2RC ECM and a multi-node heat
transfer model based on the battery geometry was employed in the study to obtain a more
detailed temperature gradient inside the large prismatic LIB. The research showed that the
hybrid model could provide similar results to the finite element method (FEM), however,
the computational burden was reduced by around 90%. They also revealed that the cell
66
Energies 2021, 14, 5960
geometry has a strong influence on the cell temperature profile. Despite good accuracy, the
effect of cell ageing and the effort of developing pack-level thermal modelling were not
considered in this study.
The impact of heat dissipation through radiation from the surface of the cell was
introduced in the thermal modelling of LIB by Sun et al. [82]. A lumped thermal model
considering the radiation effect was then used for core temperature estimation with the
help of an Extended Unscented Kalman Filter (EUKF). The sensor bias was augmented as
an extended state to enhance the prediction accuracy and model robustness. While the
load profile of residential energy storage was tested, the suitability in commercial vehicle
applications was not tested. Further, model parameters were assumed to be constant
irrespective of environmental uncertainties which may be in conflict with the facts when
the operating conditions will vary significantly. Zhu et al. [83] developed a lumped two-
state thermal-electrical model for estimating both the surface and the core temperatures
where the thermal impact of the adjacent cell was also considered during modelling.
Further, an extended state observer (ESO) with the feedback of the surface temperature was
employed to address the model uncertainties and time-variant parameters in the estimation
model. This approach is specifically designed for rapid self-heating of self-heating batteries.
The concept of model-based virtual thermal sensors (VTS) was introduced by Xiao Y. [84]
that combines the tuned thermal model with a KF observer along with an online parameter-
identification algorithm for surface and core temperature estimation utilizing a single
temperature sensor input. While the strategy is adaptive to environmental uncertainties,
it still requires a sensor for feedback; thus, the strategy cannot be termed as completely
sensorless. Despite that it minimizes the sensor requirement and enhances the model
adaptability, the concept is similar to other lumped parameter EECM-based methods.
The effect of fast-discharge on core temperature of LIB was demonstrated by Surya and
Mn [14] where a combination of 1-RC ECM, single-state thermal model and KF was used
for core temperate estimation. They used a recursive least square (RLS) algorithm to
identify model thermal parameters. However, further research is recommended to develop
health-conscious BMS suitable for fast charging/discharging.
4.2.2. EECM-Based Temperature Estimation of LIB Pack

Most of the research studies covered only the temperature estimation of a single cell.
Thermal modelling and temperature estimation of a LIB pack were seldomly reported.
A ROM of a LIB pack considering the characteristic of the inner electrical resistance of
the battery was used for core temperature estimation by Ma et al. [85]. Here, RLS was
used for the thermal parameter identification. In this study, several assumptions were
made while establishing ROM of a battery pack such that parameters of each cell are the
same and the thermal behavior of each cell row is same. The heat transfer among cells
via conduction through tabs and wires were neglected which could give rise to the error
in temperature estimation. Thermal modelling of a LIB pack by scaling-up a single cell
thermal model was investigated by Ismail et al. [86] using a simulation study. Considerable
accuracy has been noticed, however, several assumptions were made to scale up the
single-cell model to battery pack models, such as uniform cell characteristics, constant
ambient conditions and 100% efficient discharging process that are far from the real-life
scenario. Therefore, the accuracy of the temperature estimation strategy in real-world
applications needs to be further explored. Therefore, from the above discussion, it can be
stated that the pack-level estimation schemes need significant further research. The EECM-
based temperature estimation strategies proposed by different authors are summarized in
Table 2 for a quick reference to the readers. One of the major limitations of EECM-based
temperature estimation techniques is the requirement of online sensor feedback. This
is because the estimation accuracy is completely relying on accuracy of the knowledge
of the cell thermal properties, heat generation rates, and thermal boundary conditions
represented in terms of electrical parameters that are subjected to change due to cell aging,
operating temperature and other practical uncertainties.
67
Energies 2021, 14, 5960
Table 2. Summary of EECM-based temperature estimation strategies.

ANSYS (Ansys, Inc., Canonsburg, PA, USA) FLUENT
Lumped Parameter heat
Mahamud et al. [70] Simulation, validated using Experimental and
capacitance–resistance thermal model
simulation results
Lumped Parameter, Single-State,
Forgeze et al. [43] Entropy changes are considered, experimentally validated
First-order model
Lumped Parameter, Two-State,
Surya et al. [13] SOC, Surface temperature variation, MATLAB Simulation
Second-order model + Kalman Filter (KF)
High current rate, varying temperature, SOC,
Lumped Parameter, Two-State,
Lin et al. [74] experimental validation using electrochemical impedance
Second-order model
spectroscopy data
Extended lumped parameter, Two-state, Temperature variation, cell ageing, SOC, Heat transfer
Li and Yang [75] Second-order model + Forgetting factor modes, ANSYS Multiphysics Simulation, validated using
Recursive Least Square (FFRLS) experimental and simulation results
Lumped parameter, Two-state model +
Cell ageing and uncertainties in practical operation,
Lin et al. [45,76] Least square (LS) algorithm +
validated using experimental and simulation results
Nonuniform forgetting factors (NUFF)
Lumped-parameter model + Influence of overpotential entropy changes, validated
Lin et al. [77]
Closed-loop observer using a simulation study
Influence of entropy changes and overpotential, surface
Lumped parameter, Second-order, Single and ambient temperature variation, charge/discharge
Sun et al. [78]
state thermal model + KF current profile, MATLAB simulation and
experimental validation
Lumped parameter, Second-order, Initialization error and the possible time-varying external
Dai et al. [79]
Two-state model + JKF + LS algorithm thermal resistance, validated using experimental data
Doughty et al. [80] and Lumped parameter, Two-state model + Ambient temperature variation, SOC, validated using a
Park et al. [71] Extended KF simulation study
Lumped parameter, Two-state thermal Constantly varying temperature, SOC, Surface thermal
Chen et al. [72]
model + Joint KF (JKF) resistance, experimentally validated
Lumped Parameter, Second-order,
Battery geometry, charge/discharge profile, Comparison
Pan et al. [19] multi-node model +
with an FEA model, experimentally validated
particle-swarm algorithm
One-dimensional (radial) lumped Anisotropy of thermal conductivity, SOC, external
Xie et al. [81] parameter, Three node model + Dual temperature, FEM and Computational Fluid Dynamics
KF (DKF). (CFD), experimental validation
Lumped parameter, single-state model + Sensor bias, Considered heat radiation from the surface,
Sun et al. [82]
Extended unscented KF (EUKF) MATLAB simulation and experimental validation
Thermal impact of an adjacent cell, Model uncertainties
Lumped parameter, Two-state model + and time-variant parameters, MATLAB simulation,
Zhu et al. [83]
extended state observer (ESO) validation and comparison using electrochemical
impedance spectroscopy data
Lumped parameter, Single-state thermal
Surya and Mn [14] model + KF + Recursive Least Square Effect of fast-discharge, MATLAB Simulation
(RLS) algorithm
EECM-based virtual thermal sensors Environmental uncertainties were considered, validated
Xiao Y. [84]
(VTS) + KF using experimental and simulation results
ROM of a LIB pack for a central
Ma et al. [85] and Temperature, SOC, validated using experimental and
temperature of LIB pack + Recursive least
Ismail et al. [86] simulation results
square (RLS)
68
Energies 2021, 14, 5960
4.3. Numerical Analysis-Based Temperature Estimation

Numerical method-based techniques were successfully implemented for tempera-
ture estimation of different chemistries and shapes of LIB cells and even LIB packs. So
far, the finite element method (FEM) [87–90] and finite volume method (FVM) [91] were
extensively used for temperature estimation. Numerical analysis-based techniques try to
mathematically describe the thermal dynamics inside the battery using nonlinear partial
differential equations (PDEs) such as used by Du et al. [89]. They have employed FEM anal-
ysis with a three-dimensional model and Bernardi equation-based internal heat generation
rate. Typically, the PDEs have complex boundary conditions that are infinite-dimensional.
The fundamental mathematical equation as employed by Du et al. can be represented as
Equation (5)
∂T ∂2 T δ2 T δ2 T
ρC p = λ x 2 + λy 2 + λz 2 + Q (5)
∂t ∂x δy δz
where ρ, C p represent the mean density and mean specific heat of the cell, respectively. λ is
the heat conductivity coefficient of the surface material of the cell and Q is the same as in
Equation (1).
Dong Hyup Jeon [87] incorporated a transient thermoelectric model with a porous
electrode model and conducted a numerical simulation to understand the thermal behavior
of a commercial LIB under charging and discharging conditions. He demonstrated that
temperature increase during discharging is much higher compared to the temperature rise
during charging. He also suggested that the temperature difference between charge and
discharge can be decreased with increasing C-rates. Further, Baba et al. [88] conducted
a numerical simulation of an enhanced single-particle model of a LIB to understand the
three-dimensional temperature distribution inside the cell. Numerical analysis was used
for transient behaviors of a LIB under a dynamic driving cycle by Yi et al. [90]. Double-
layer thermal capacitance was used to capture the short-term transient behavior of the LIB
chemistry. Fleckenstein et al. [91] using FVM to demonstrate that the temperature gradients
inside the cell layer result in different current densities and local SOC inhomogeneities in
LIB. These phenomena must be well-taken care of while designing an effective thermal
management system. In general, this kind of model is best for capturing both temporally
and spatially thermal distribution of the cell as the battery thermal process is a typical
distributed parameter system. Despite high accuracy and detailed information about cell
temperature gradient, these numerical method-based temperature estimation strategies are
not suitable for online temperature estimation due to high computational cost. The complex
mathematical analysis also required expertise and strong domain knowledge. Moreover,
generalization is not possible as different chemistry and cell physics affect mathematical
modelling. A summary of numerical methods-based temperature estimation strategies is
shown in Table 3.
Table 3. Summary of numerical methods-based temperature estimation strategies.

A transient thermoelectric
model with a porous electrode Different driving cycles, COMSOL Multiphysics (COMSOL Inc.,
Dong Hyup Jeon [87]
model + finite element Stockholm, Sweden) simulation and Experimental validation
method (FEM)
Enhanced single-particle Three-dimensional temperature distribution inside the cell, cell
Baba et al. [88]
model + FEM geometry, and current profile, experimentally validated
Three-dimensional model +
Different current profiles, temperature variation, COMSOL
Du et al. [89] ECM based heat generation
Multiphysics simulation and experimental validation
model + FEM
69
Energies 2021, 14, 5960
Table 3. Cont.

Transient thermoelectric Transient behaviors under dynamic driving cycle
Yi et al. [90]
model + FEM Experimentally validated
Different current density and local SOC inhomogeneities at
Three-dimensional
Fleckenstein et al. [91] different cell layers, MATLAB (MathWorks, MA, USA) simulation
model + FVM
and experimental validation
4.4. Direct Impedance Measurement-Based Temperature Estimation

Cell internal temperature estimation using a lumped-parameter thermal model and an
approximate distributed thermal model have several drawbacks. Firstly, accurate determi-
nation of thermal model parameters such as heat generation and cell thermal properties is
highly challenging. Heat generation inside the cell is typically approximated by measuring
the cell operating current, voltage and the internal resistance that are again functions of
SOC, cell internal temperature and SOH. Moreover, a cell is constructed using many differ-
ent materials combined into a layered structure and thermal contact resistances between
these layers are often unknown. Temperature estimation methods use surface temperature
measurements and even the combination of surface-mounted temperature sensor and
thermal model typically failed to detect the thermal runaway as rapid fluctuations in the
internal temperature is difficult to capture using surface mounted sensors because the
heat conduction between the core and battery surface takes a considerable amount of
time [92]. Furthermore, embedding micro-temperature sensors within the cell [93,94] is
not practically possible for a large capacity LIB pack from a manufacturing complexity
and system cost point of view. Hence, the core temperature measurement using a physical
sensor is not an appropriate method for industrial applications.
Srinivasan et al. [95,96] noticed that the phase of electrochemical impedance in the
frequency range of 40 to 100 Hz is temperature-sensitive but insensitive to changes in
other parameters such as SOC and SOH. Based on these findings, they demonstrated an
electrochemical impedance-based cell internal temperature estimation strategy. However,
they assumed the uniform internal temperature and the estimation method is only valid
in the temperature range of from −20 to 66 ◦ C. The temperature estimation considering
the effect of temperature non-uniformity on electrochemical impedance was studied by
Schmidt et al. [97] based on the principle derived by Troxler et al. [98]. Both the strategy
developed by Srinivasan et al. and Schmidt et al. were only able to estimate the mean
temperature of the cell, however, in real-life application, especially in the case of cylin-
drical battery under high charging/discharging current, the difference between internal
maximum temperature, surface temperature and mean temperature are significantly high.
Therefore, Richardson et al. [3] further extended the research and developed a thermal-
impedance model by combining an EIS measurement at a single frequency with a surface
temperature measurement for precise determination of internal temperature distribution.
The fundamental steps in direct impedance measurement-based temperature estimation as
presented by Richardson et al. [3] is shown in Figure 4.
The approach of Richardson et al. does not require knowledge of cell thermal prop-
erties, heat generation or thermal boundary conditions, however, the major limitation
is the online impedance determination of each cell which is highly challenging. More-
over, uncertainties of environmental factors were not considered and a surface-mounted
temperature sensor needs to be installed on each cell which is impractical so far. Whilst
few approaches of online determination of impedance spectra across multiple frequen-
cies using onboard power electronics of EVs have been reported [99], the application of
these strategies in real-time temperature estimation has not yet been investigated. Further-
more, interpreting impedance measurements under superimposed DC currents is yet to be
systematically investigated.
70
Energies 2021, 14, 5960
Figure 4. Fundamental steps in direct impedance measurement-based temperature estimation.
Online EIS-based temperature estimation strategy termed impedance-temperature

detection (ITD) was proposed by Richardson and Howey [100] for sensorless temperature
estimation which is adaptive to cell ageing and practical uncertainties. However, ITD
cannot provide a general solution alone, thus, such a strategy combines surface-mounted
sensors with ITD for accurate online temperature estimation [3]. Still, temperature sensors
are required to be installed. Further to this study, they integrated ITD with an electric-
thermal model along with a DEKF for online core temperature estimation of a LIB cell even
with unknown convection coefficient. They also demonstrated that the performance of the
thermal model plus ITD is almost similar to the ITD with surface thermal sensors. Despite
the advantages, the major limitations of the strategy are online impedance determination
and the requirement of an accurate electric thermal model, thus encompassing the same
drawback of conventional thermal modelling-based strategies. Moreover, although the
strategy can estimate both core and surface temperature of an individual cell, the pack-level
estimation strategy was not illustrated in this study.
The influence of cell temperature, SOC and SOH on the impedance spectrum, ex-
citation frequency and thereby estimation accuracy of cell internal temperature was in-
vestigated by Zhu et al. [101]. Here, the temperature estimation was made based on
an impedance response matrix analysis which was developed using EIS measurements.
Despite high accuracy, the effect of the nonuniformity of the cell temperature and the cor-
rection method was not considered. Moreover, an extensive experimental study is required
for modelling and the computational cost is also very high. Thus, the online application of
the strategy is challenging. Identification of suitable frequency and other EIS parameters
is very difficult whilst the estimation accuracy significantly depends on these parameters.
Moreover, accurate determination of the real and imaginary parts of the impedance is
highly challenging, whilst different decisions for these two parts leads to inaccurate tem-
71
Energies 2021, 14, 5960
perature estimation. A combination of Linear Parameter Varying (LPV) thermal model and
a polytopic observer-based battery-cell temperature estimation algorithm was proposed
by Debert et al. [102]. The EIS-based strategy was also employed in references [3,103–106]
to estimate the core temperature. Despite high accuracy, the major limitation is the de-
termination of accurate impedance-temperature characteristics and it should be acquired
in advance through tedious preliminary tests. In addition, the impedance-temperature
characteristic of a cell is influenced by cell ageing leading to inaccurate prediction due
to SOH deterioration. A summary of direct impedance measurement-based temperature
estimation strategies is presented in Table 4.
Table 4. Summary of direct impedance measurement-based strategies.

Direct measurement of
Srinivasan et al. [95,96] Experimental validation with EIS data
electrochemical impedance
Direct measurement of Temperature non-uniformity was not considered,
Schmidt et al. [97]
electrochemical impedance experimentally validated
Thermal-impedance model + EIS Independent of cell thermal properties, heat
Richardson et al. [3] measurement at single frequency + generation or thermal boundary conditions,
surface temperature feedback experimental validation with EIS data
Online EIS measurement
Unknown convection coefficient is considered,
Richardson and Howey [100] (impedance-temperature detection (ITD)
experimentally validated
+ dual-extended Kalman filter (DEKF)
Influence of cell temperature, SOC and SOH on the
Impedance response matrix analysis,
Zhu et al. [101] impedance spectrum, experimental validation with
developed using EIS measurements
EIS data
4.5. Machine Learning-Based Temperature Estimation

With the overwhelming complexity of the electrochemical reactions inside the battery
and the sensitivity of the battery parameters to the uncertainties of the working environ-
ment, the thermodynamic behavior varies significantly from the center region to the surface
region. Most of the existing distributed thermal models and the lumped parameter thermal
models are incapable to consider the spatiotemporal distribution of LIB packs, especially
in the case of large-capacity battery packs. Moreover, it is highly difficult to represent these
spatiotemporal dynamics by a single physics-based model. Here, the machine learning
(ML) algorithms were widely employed to preserve the local dynamics to improve the
modelling accuracy of nonlinear systems such as LIB. A schematic layout of the ML-based
temperature estimation scheme is shown in Figure 5.
Liu and Li [107] employed a hybrid model of EECM and neural network (NN)-
based learning approach to develop a spatiotemporal thermodynamic model of LIB for
accurate estimation of internal temperature distribution. The data-driven NN model used
commonly measured signals of BMS to compensate for the model-plant mismatch caused
by spatial nonlinearity and other model uncertainties. NN and support vector machine
(SVM)-based [108] LIB temperature estimation strategy was investigated by Sbarufatti
et al. [109]. A hybrid model of an radial basis function neural network (RBFNN) and EKF
was employed by Liu et al. [110] to estimate the internal temperature of LIB. While they
have considered the impact of temperature on cell behavior, the primary intention of these
models was the estimation of SOC or SOH rather than estimating the cell temperature.
One of the major challenges of pure ML-based strategies is the generalization capability.
Feng et al. [111] developed an effective electrochemical-thermal-neural-network (ETNN) by
fusing a lumped parameter electrochemical thermal, feed-forward neural network (FFNN)
and a UKF. This method demonstrated appreciable performance in predicting the state of
temperature (SOT) in a wide temperature range and large current conditions. However, the
modelling is highly complex, the accuracy over different charging current/drive cycles was
72
Energies 2021, 14, 5960
not tested. Moreover, the computational efficiency and the suitability for online application
are questionable. The back of the ETNN is the electrochemical model thus encompassing
drawbacks similar to electrochemical models. In general, while ML-based schemes are
computationally efficient, collecting training data and model training procedures are highly
complex and time expensive. Moreover, real-life battery test data were not considered
during ML-based model training in the existing literature; therefore, the accuracy of the
existing ML-based strategies is still questionable. A summary of ML-based techniques
reported by researchers is presented in Table 5.
Figure 5. Schematic layout of ML-based temperature estimation scheme.
Table 5. Summary of ML-based temperature estimation techniques.

Model-plant mismatch caused by spatial nonlinearity and other
EECM + neural network
Liu and Li [107] model uncertainties, NN-model was validated using
(NN)-based learning approach
experimental data
Influence of temperature, charging/discharging current, Python
Neural networks + Support
Sbarufatti et al. [109] (Python Software Foundation, Wilmington, DE, USA), NN and
vector machines
SVM-model were validated using experimental data
73
Energies 2021, 14, 5960
Table 5. Cont.

RBF neural network (RBNN) and Impact of temperature on cell behavior, validated using
Liu et al. [110]
the extended Kalman filter (EKF) simulation data
Electrochemical-thermal-neural-
Wide temperature and large current conditions, Python, validated
Feng et al. [111] network (ETNN) + Unscented
using experimental data
Kalman filter (UKF).
4.6. Magnetic Nanoparticles-Based Temperature Estimation

The magnetization of Magnetic Nanoparticles (MNPs) is nonlinear under an ac mag-
netic field and the accurate temperature of MNPs could be estimated by using the ratio
of the third and fifth harmonic response [112–114]. Further, the temperature sensitivity of
MNPs with the increased DC magnetic field was studied by Zhong et al. [115]. They found
that the temperature sensitivity of MNPs will decrease with an increased DC magnetic
field. Further to this study, Zou et al. [116] developed an improved Magnetic nanoparticles
thermometer (MNPT) for the core temperature estimation of LIB which works based on
the temperature measurement of magnetic nanoparticles (MNPs). They also suggested
the optimal range of the DC magnetic field strength to ensure maximum temperature
sensitivity and minimum temperature error of the MNPT. It is noticed that this type of
estimation topology is very bulky and costly. Moreover, the suitability of online prediction
has not yet been assessed.
5. Discussion on Issues, Challenges and Future Research Recommendations

Temperature estimation schemes for LIBs can be designed with different levels of
complexity depending on the requirement of accuracy level and detailing of the prediction
results. Detailed model results and more accurate predictions are essential for safer and
reliable operation of BMS. However, integrating more detailed cell phenomena into the
model eventually increases the modelling complexity, computational cost while, at the same
time, reduces the suitability for online prediction and low-cost onboard BMS. For instance,
modelling complexity increases if the temperature gradient of each cell layer is considered
instead of concentrated heat generation at the core. Secondly, the heat fluxes inside
and outside the battery can be considered in both axial and radial directions instead of
considering only the radial direction for simplicity. Furthermore, detailed models typically
consider different heat transport modes, that is, conduction, convective and radiation
whereas simplified models consider only conduction heat transfer. Integrating a greater
number of phenomena in thermal modelling requires a lot of parameters, resulting in
additional requirements of experimental measurements, modelling time and solid domain
knowledge. In addition, very detailed and accurate information of cell structure, material
properties and cell assembly are also needed. However, collecting this information from
the cell manufacturer is highly challenging due to the confidentiality of the design data.
Therefore, it can be inferred from the above discussion that the detailed models could
produce highly accurate and complete insight into cell thermodynamics, however, their
computational complexity may not be suitable for online prediction and onboard low-cost
BMS. In general, most of the estimation strategies require measurements from physical
sensors, however, installing a physical sensor at each cell is not practically possible as a high-
capacity LIB pack consists of thousands of individual cells. Moreover, installing a sensor at
the cell core for core temperature measurement is highly challenging. Several estimation
schemes estimate the core temperature based on the surface temperature measurement.
However, it is very erroneous as it takes a significant amount of time for the heat to reach
the surface from the core. So far, most of the research studies have covered the temperature
estimation scheme of a single LIB cell. Temperature estimation of a LIB pack is much
more challenging. Thus, significant further research is recommended here. Moreover, the
influence of fast charging/discharging on the cell temperature has not yet been deeply
74
Energies 2021, 14, 5960
explored. It is highly recommended to develop a health-conscious BMS. A summary of

existing issues, challenges and future research recommendations to the research community
are presented in Table 6.
Table 6. Summary of major issues, challenges and research recommendations.
Strategy Major Issues and Challenges Future Research Recommendations
• Extremely detailed modelling is possible. Thus, it could

• Significant future research is
produce a highly accurate prediction, however with the
recommended to reduce modelling
expenses of very high computational cost. Thus, unsuitable
complexity and computational cost
for online prediction by onboard BMS
• So far, it can produce the best prediction
• In-depth prior knowledge of LIB chemistry is a must besides
results, thus could be extensively used
Electrochemical expertise in mathematical modelling, resulting in dependence
for the validation of other types of
Model-based on domain experts
models and data acquisition for
• Extensive experiments are required to accumulate detailed
data-driven models
information on battery characteristics
• LIB chemistry is highly sensitive to
• Modelling is highly complex
temperature, battery health and other
• Developing an adaptive estimation scheme is
uncertainties, thus, further research on
highly challenging
adaptive modelling is recommended
• Poor generalization capability
• Modelling complexity and

Computational cost can be controlled by
• Most extensively used so far due to adequate accuracy and treading-off between accuracy
easy implementation, however, modelling complexity and requirement and detailing of the model
computational cost increase with the order of the model, • Adaptive parameterization is
number of temperature measurement points (nodes) and the challenging, however with the fusion of
parameter distribution advanced algorithms such as ML-based
• Accurate EECM parameters are very difficult to identify, techniques, adaptive strategies could be
Equivalent Electric especially online parameter estimation developed
Circuit (EECM) • Parameter tuning using external measurement is challenging • These models can generate highly
Model-based and time expensive accurate results at the laboratory, thus
• Few researchers also used electrochemical analysis for could be used to generate data and
parameter identification and determination which possesses model validation of other strategies
similar difficulties to electrochemical-based strategies • Fusion of this strategy with other
• Predictions are highly influenced by measurement noises and strategies such as ML-based techniques
often too many physical sensors are required could produce enhanced accuracy and
• Lower order models/simplified models are so far extensively computational performance
used for online prediction with the compromise of accuracy • Instead of traditional filters, more
and detailed insight advanced adaptive filtering techniques
could be embedded for better
performance
• Completely data-driven black-box strategy, that is, prediction

depends on the external measurements only, thus, minimal or • While it is comparatively easy to develop
no requirement of any domain-specific knowledge, however, adaptive models, however, very few
the major challenge is the accumulation of high-quality large efforts have been made so far.
volume of training data • Cell characteristics are highly influenced
• No requirement of iterative complex mathematical by temperature, ageing and other
calculation, thus, computational cost is adequate for online uncertainties, thus, further research on
application, however, computational cost increases with the adaptive modelling is recommended
high volume (high resolution) data and number of feature • Generalization is difficult, however, with
Machine Learning vectors to obtain a better insight the incorporation of advanced adaptive
(ML)-based • Accumulation of high-resolution data especially algorithms, it could be possible
manufacturer data and fault data are highly challenging. • With proper design efforts, it could be
These data are important for accurate and adaptive prediction used for online prediction and
• Generalization is challenging implemented in onboard BMS with low
• Currently not used on onboard BMS due to high training time processing power
and complex algorithm development and computational time, • Very promising technology could be
whilst it is noticed that very few efforts have been made so far used for a future generation of sensorless
• Often, external measurements by physical sensors are temperature estimation strategies. Very
required as feedback for online parameter adjustment, thus little effort has been provided so far, thus,
still requires installation of physical sensors further research is recommended
75
Energies 2021, 14, 5960
Table 6. Cont.
Strategy Major Issues and Challenges Future Research Recommendations
• These strategies use FEA and FVA. FEA and FVA based • Significant research and development are
temperature estimation strategies are considered the most required to improve computational cost
accurate and most computationally expensive to make this suitable for online
Numerical Model-based
• Due to iterative complex mathematical calculation, its prediction.As it is most accurate, it could
computational cost is very high, thus not suitable for be used for other model validation and
online prediction accurate data collection
• Promising technology, thus, significant

• The influence of temperature on cell impedance is used for further research and development is
internal temperature estimation. However, online direct recommended to reduce scheme size and
measurement of impedance using onboard power electronics assess the practical applicability in
is highly challenging onboard BMS
Direct Impedance • Changes in cell impedance due to temperature variation is • Accuracy in real-world applications
Measurement-based small, thus, accurate determination of such small changes is needs to be judged
highly difficult • Further research into online impedance
• Existing schemes are very bulky determination using onboard electronics
• Very few research efforts have so far provided, not yet is also recommended
practically implemented • The cost of existing solutions is very
high, which needs to be addressed
• Practical applicability in onboard

Magnetic low-cost BMS has not yet been
Nanoparticle-based • Very new technology, it is too early to comment investigated. Overall, significant further
research is required
6. Conclusions
This article presented a comprehensive review of the state-of-the-art temperature
estimation strategies for lithium-ion batteries (LIBs) covering the necessity of an optimum
estimation strategy, detailed discussion on the existing strategies, current issues, challenges
and future research recommendations. It can be inferred that an accurate temperature
estimation of LIBs is indispensable for effective thermal management, operational safety
and several other crucial tasks of a Battery Management System (BMS). Measurement of
each cell temperature using physical sensors is not practically possible, especially for a
high-capacity battery pack consisting of thousands of individual cells. To develop an ideal
temperature estimation scheme, one needs to concentrate on several factors, such as high
accuracy, adaptability, small size, real-time estimation, distribution (to monitor the tem-
perature gradient of the entire cell), low cost, and easily implementable for wide adoption.
Typically, a temperature estimation scheme consists of a heat generation model and a heat
transfer model. Depending on the modelling and computation strategies temperature esti-
mation schemes can be grouped into six categories, namely, electrochemical model-based,
equivalent electric circuit model (EECM)-based, machine learning (ML)-based, numerical
analysis-based, direct impedance measurement-based, and magnetic nanoparticle-based.
So far, numerical analysis-based schemes are most accurate followed by electro-chemical
model-based schemes. However, both strategies have very high computational cost making
them inappropriate for online prediction by a low-cost onboard BMS. Moreover, mod-
elling complexity and experimental requirements are very high alongside the necessity of
domain-specific knowledge. EECM-based schemes can be designed with different levels of
complexity, accuracy level and computational cost. Simplified lower-order EECM-based
schemes are extensively used in the literature and practice. Machine learning (ML)-based
schemes are very promising due to their higher level of accuracy, ease of implementation
and adaptability. In addition, reduced or even no requirement of equivalent modelling and
domain experts. However, to obtain the feature vectors, very large volume and high-quality
data are required which are typically very challenging to acquire. Here, a hybrid strategy
combining an EECM and an ML is presumably a suitable solution. Direct impedance
measurement and magnetic nanoparticle-based schemes are very newly developed. It is
too early to assess their capability and suitability for online prediction and implementation
in onboard BMS. Therefore, systematic guidelines about open research areas and future
76
Energies 2021, 14, 5960
research directions are highlighted in this study. It is also noticed that the majority of the
research studies proposed temperature estimation schemes of a single LIB cell whereas
temperature estimation of a LIB pack is much more challenging. Thus, significant further
research is recommended here a well.
Author Contributions: Conceptualization, S.S.W.; methodology, A.S. and S.S.W.; investigation, A.S.;
resources, S.S.W.; writing—original draft preparation, A.S.; writing—review and editing, A.S. and
S.S.W.; visualization, A.S.; supervision, S.S.W. All authors have read and agreed to the published
version of the manuscript.
Funding: This research received no external funding.
Data Availability Statement: Not applicable.
References
1. Jilte, R.D.; Kumar, R.; Ma, L. Thermal performance of a novel confined flow Li-ion battery module. Appl. Therm. Eng. 2019, 146,
1–11. [CrossRef]
2. Yang, X.-H.; Tan, S.-C.; Liu, J. Thermal management of li-ion battery with liquid metal. Energy Convers. Manag. 2016, 117, 577–585.
[CrossRef]
3. Richardson, R.R.; Ireland, P.T.; Howey, D. Battery internal temperature estimation by combined impedance and surface tempera-
ture measurement. J. Power Sources 2014, 265, 254–261. [CrossRef]
4. Wang, Y.-F.; Wu, J.-T. Performance improvement of thermal management system of lithium-ion battery module on purely electric
AUVs. Appl. Therm. Eng. 2019, 146, 74–84. [CrossRef]
5. Surya, S.; Samanta, A.; Marcis, V.; Williamson, S.S. Smart core and surface temperature estimation techniques for health-conscious
lithium-ion battery management systems: A model-to-model comparison. Preprints 2021, 1–21. [CrossRef]
6. Tanim, T.R.; Rahn, C.D.; Wang, C.-Y. State of charge estimation of a lithium ion cell based on a temperature dependent and
electrolyte enhanced single particle model. Energy 2015, 80, 731–739. [CrossRef]
7. Farmann, A.; Sauer, D.U. A study on the dependency of the open-circuit voltage on temperature and actual aging state of
lithium-ion batteries. J. Power Sources 2017, 347, 1–13. [CrossRef]
8. Zheng, F.; Jiang, J.; Sun, B.; Zhang, W.; Pecht, M. Temperature dependent power capability estimation of lithium-ion batteries for
hybrid electric vehicles. Energy 2016, 113, 64–75. [CrossRef]
9. Samanta, A.; Chowdhuri, S. Active cell balancing of lithium-ion battery pack using dual dc-dc converter and auxiliary lead-acid
battery. J. Energy Storage 2021, 33, 102109. [CrossRef]
10. Samanta, A.; Chowdhuri, S.; Williamson, S. Machine learning-based data-driven fault detection/diagnosis of lithium-ion battery:
A critical review. Electronics 2021, 10, 1309. [CrossRef]
11. Zhang, S.; Xu, K.; Jow, T. The low temperature performance of li-ion batteries. J. Power Sources 2003, 115, 137–140. [CrossRef]
12. Mohan, S.; Kim, Y.; Stefanopoulou, A.G. Energy-conscious warm-up of li-ion cells from subzero temperatures. IEEE Trans. Ind. Electron.
2016, 63, 2954–2964. [CrossRef]
13. Surya, S.; Marcis, V.; Williamson, S. Core temperature estimation for a lithium ion 18650 cell. Energies 2020, 14, 87. [CrossRef]
14. Surya, S.; Mn, A. Effect of fast discharge of a battery on its core temperature. In Proceedings of the 2020 International Conference
on Futuristic Technologies in Control Systems & Renewable Energy (ICFCR), Kerala, India, 23–24 September 2020. [CrossRef]
15. Wei, Z.; Zhao, J.; He, H.; Ding, G.; Cui, H.; Liu, L. Future smart battery and management: Advanced sensing from external to
embedded multi-dimensional measurement. J. Power Sources 2021, 489, 229462. [CrossRef]
16. Steinhorst, S.; Lukasiewycz, M.; Narayanaswamy, S.; Kauer, M.; Chakraborty, S. Smart cells for embedded battery management.
In Proceedings of the 2014 IEEE International Conference on Cyber-Physical Systems, Networks, and Applications, Hong Kong,
China, 25–26 August 2014.
17. Huang, X.; Sui, X.; Stroe, D.-I.; Teodorescu, R. A review of management architectures and balancing strategies in smart batteries.
In Proceedings of the IECON 2019—45th Annual Conference of the IEEE Industrial Electronics Society, Lisbon, Portugal,
14–17 October 2019.
18. Fleming, J.; Amietszajew, T.; Charmet, J.; Roberts, A.J.; Greenwood, D.; Bhagat, R. The design and impact of In Situ and operando
thermal sensing for smart energy storage. J. Energy Storage 2019, 22, 36–43. [CrossRef]
19. Pan, Y.-W.; Hua, Y.; Zhou, S.; He, R.; Zhang, Y.; Yang, S.; Liu, X.; Lian, Y.; Yan, X.; Wu, B. A computational multi-node
electro-thermal model for large prismatic lithium-ion batteries. J. Power Sources 2020, 459, 228070. [CrossRef]
20. Bernardi, D.M.; Pawlikowski, E.M.; Newman, J. A general energy balance for battery systems. J. Electrochem. Soc. 1985, 132, 5–12.
[CrossRef]
77
Energies 2021, 14, 5960
21. Sun, F.; Xiong, R.; He, H.; Li, W.; Aussems, J.E.E. Model-based dynamic multi-parameter method for peak power estimation of
22. Lu, L.; Han, X.; Li, J.; Hua, J.; Ouyang, M. A review on the key issues for lithium-ion battery management in electric vehicles.
23. Ghalkhani, M.; Bahiraei, F.; Nazri, G.-A.; Saif, M. Electrochemical-thermal model of pouch-type lithium-ion batteries.
Electrochim. Acta 2017, 247, 569–587. [CrossRef]
24. Yang, X.-G.; Leng, Y.; Zhang, G.; Ge, S.; Wang, C.-Y. Modeling of lithium plating induced aging of lithium-ion batteries: Transition
from linear to nonlinear aging. J. Power Sources 2017, 360, 28–40. [CrossRef]
25. Chen, M.; Bai, F.; Song, W.; Lv, J.; Lin, S.; Feng, Z.; Li, Y.; Ding, Y. A multilayer electro-thermal model of pouch battery during
normal discharge and internal short circuit process. Appl. Therm. Eng. 2017, 120, 506–516. [CrossRef]
26. Zhao, Y.; Diaz, L.B.; Patel, Y.; Zhang, T.; Offer, G.J. How to cool lithium ion batteries: Optimising cell design using a thermally
coupled model. J. Electrochem. Soc. 2019, 166, A2849–A2859. [CrossRef]
27. Damay, N.; Forgez, C.; Bichat, M.-P.; Friedrich, G. Thermal modeling of large prismatic LiFePO4 /graphite battery. Coupled
thermal and heat generation models for characterization and simulation. J. Power Sources 2017, 283, 37–45. [CrossRef]
28. Allafi, W.; Zhang, C.; Uddin, K.; Worwood, D.; Dinh, T.Q.; Ormeno, P.A.; Li, K.; Marco, J. A lumped thermal model of lithium-ion
battery cells considering radiative heat transfer. Appl. Therm. Eng. 2018, 143, 472–481. [CrossRef]
29. Esmaeili, J.; Jannesari, H. Developing heat source term including heat generation at rest condition for lithium-ion battery pack by
up scaling information from cell scale. Energy Convers. Manag. 2017, 139, 194–205. [CrossRef]
30. Arora, S.; Shen, W.; Kapoor, A. Neural network based computational model for estimation of heat generation in LiFePO4 pouch
cells of different nominal capacities. Comput. Chem. Eng. 2017, 101, 81–94. [CrossRef]
31. Hu, X.; Liu, W.; Lin, X.; Xie, Y. A comparative study of control-oriented thermal models for cylindrical li-ion batteries.
IEEE Trans. Transp. Electrif. 2019, 5, 1237–1253. [CrossRef]
32. Xie, Y.; Li, W.; Hu, X.; Zou, C.; Feng, F.; Tang, X. Novel mesoscale electrothermal modeling for lithium-ion batteries.
IEEE Trans. Power Electron. 2019, 35, 2595–2614. [CrossRef]
33. Friesen, A.; Mönnighoff, X.; Börner, M.; Haetge, J.; Schappacher, F.M.; Winter, M. Influence of temperature on the aging behavior
of 18650-type lithium ion cells: A comprehensive approach combining electrochemical characterization and post-mortem analysis.
34. Fan, Y.; Bao, Y.; Ling, C.; Chu, Y.; Tan, X.; Yang, S. Experimental study on the thermal management performance of air cooling for
high energy density cylindrical lithium-ion batteries. Appl. Therm. Eng. 2019, 155, 96–109. [CrossRef]
35. Saw, L.H.; Poon, H.M.; Thiam, H.S.; Cai, Z.; Chong, W.T.; Pambudi, N.A.; King, Y.J. Novel thermal management system using
mist cooling for lithium-ion battery packs. Appl. Energy 2018, 223, 146–158. [CrossRef]
36. Liu, B.; Yin, S.; Xu, J. Integrated computation model of lithium-ion battery subject to nail penetration. Appl. Energy 2016, 183,
37. Doyle, M.; Fuller, T.F.; Newman, J.S. Modeling of galvanostatic charge and discharge of the lithium/polymer/insertion cell.
J. Electrochem. Soc. 1993, 140, 1526–1533. [CrossRef]
38. Xiao, Y.; Fahimi, B. State-space based multi-nodes thermal model for lithium-ion battery. In Proceedings of the 2014 IEEE
Transportation Electrification Conference and Expo (ITEC), Beijing, China, 31 August 2014.
39. Tian, N.; Fang, H.; Wang, Y. 3-D Temperature field reconstruction for a lithium-ion battery pack: A distributed kalman filtering
approach. IEEE Trans. Control. Syst. Technol. 2019, 27, 847–854. [CrossRef]
40. Ruan, H.; Jiang, J.; Sun, B.; Gao, W.; Wang, L.; Zhang, W. Online estimation of thermal parameters based on a reduced
wide-temperature-range electro-thermal coupled model for lithium-ion batteries. J. Power Sources 2018, 396, 715–724. [CrossRef]
41. Gu, W.B.; Wang, C.Y. Thermal-electrochemical modeling of battery systems. J. Electrochem. Soc. 2000, 147, 2910–2922. [CrossRef]
42. Kumaresan, K.; Sikha, G.; White, R.E. Thermal model for a li-ion cell. J. Electrochem. Soc. 2008, 155, A164–A171. [CrossRef]
43. Forgez, C.; Do, D.V.; Friedrich, G.; Morcrette, M.; Delacourt, C. Thermal modeling of a cylindrical LiFePO4 /graphite lithium-ion
battery. J. Power Sources 2010, 195, 2961–2968. [CrossRef]
44. Dees, D.W.; Battaglia, V.S.; Bélanger, A. Electrochemical modeling of lithium polymer batteries. J. Power Sources 2002, 110, 310–320.
[CrossRef]
45. Lin, X.; Perez, H.E.; Siegel, J.; Stefanopoulou, A.G.; Li, Y.; Anderson, R.D.; Ding, Y.; Castanier, M. Online parameterization of
lumped thermal dynamics in cylindrical lithium ion batteries for core temperature estimation and health monitoring. IEEE Trans.
Control Syst. Technol. 2012, 21, 1745–1755. [CrossRef]
46. Choi, J.W.; Aurbach, D. Promise and reality of post-lithium-ion batteries with high energy densities. Nat. Rev. Mater. 2016,
47. Yang, F.; Wang, D.; Zhao, Y.; Tsui, K.-L.; Bae, S.J. A study of the relationship between coulombic efficiency and capacity
degradation of commercial lithium-ion batteries. Energy 2018, 145, 486–495. [CrossRef]
48. Wang, Q.; Zhao, X.; Ye, J.; Sun, Q.; Ping, P.; Sun, J. Thermal response of lithium-ion battery during charging and discharging
under adiabatic conditions. J. Therm. Anal. Calorim. 2016, 124, 417–428. [CrossRef]
49. Chen, Y. Heat transfer phenomena in lithium/polymer-electrolyte batteries for electric vehicle application. J. Electrochem. Soc.
1993, 140, 1833–1838. [CrossRef]
78
Energies 2021, 14, 5960
50. Fuller, T.F.; Doyle, M.; Newman, J. Relaxation phenomena in lithium-ion-insertion cells. J. Electrochem. Soc. 1994, 141, 982–990.
[CrossRef]
51. Pals, C.R.; Newman, J. Thermal modeling of the lithium/polymer battery: I discharge behavior of a single cell. J. Electrochem. Soc.
1995, 142, 3274–3281. [CrossRef]
52. Pals, C.R.; Newman, J. Thermal modeling of the lithium/polymer battery: II temperature profiles in a cell stack. J. Electrochem. Soc.
1995, 142, 3282–3288. [CrossRef]
53. Al Hallaj, S.; Maleki, H.; Hong, J.; Selman, J. Thermal modeling and design considerations of lithium-ion batteries. J. Power Sources
1999, 83, 1–8. [CrossRef]
54. Gomadam, P.M.; White, R.E.; Weidner, J.W. Modeling heat conduction in spiral geometries. J. Electrochem. Soc. 2003, 150,
A1339–A1345. [CrossRef]
55. Chen, Y.; Evans, J.W. Three-dimensional thermal modeling of lithium-polymer batteries under galvanostatic discharge and
dynamic power profile. J. Electrochem. Soc. 1994, 141, 2947–2955. [CrossRef]
56. Newman, J.; Tiedemann, W. Temperature rise in a battery module with constant heat generation. J. Electrochem. Soc. 1995, 142,
1054–1057. [CrossRef]
57. Chen, Y.; Evans, J.W. Thermal analysis of lithium-ion batteries. J. Electrochem. Soc. 1996, 143, 2708–2712. [CrossRef]
58. Rao, L.; Newman, J. Heat-generation rate and general energy balance for insertion battery systems. J. Electrochem. Soc. 1997, 144,
2697–2704. [CrossRef]
59. Kim, G.-H.; Pesaran, A.; Spotnitz, R. A three-dimensional thermal abuse model for lithium-ion cells. J. Power Sources 2007, 170,
60. Thomas, K.E.; Newman, J.; Darling, R.M. Mathematical modeling of lithium batteries. In Advances in Lithium-Ion Batteries;
Springer: Berlin/Heidelberg, Germany, 2002; pp. 345–392.
61. Van Schalkwijk, W.; Scrosati, B. Advances in lithium ion batteries introduction. In Advances in Lithium-Ion Batteries; Springer:
Berlin/Heidelberg, Germany, 2002.
62. Smith, K.; Wang, C.-Y. Power and thermal characterization of a lithium-ion battery pack for hybrid-electric vehicles.
63. Khateeb, S.A.; Amiruddin, S.; Farid, M.; Selman, J.R.; Al-Hallaj, S. Thermal management of li-ion battery with phase change
material for electric scooters: Experimental validation. J. Power Sources 2005, 142, 345–353. [CrossRef]
64. Fang, W.; Kwon, O.J.; Wang, C.-Y. Electrochemical-thermal modeling of automotive li-ion batteries and experimental validation
using a three-electrode cell. Int. J. Energy Res. 2009, 34, 107–115. [CrossRef]
65. Kim, U.S.; Yi, J.; Shin, C.B.; Han, T.; Park, S. Modelling the thermal behaviour of a lithium-ion battery during charge.
66. Gerver, R.E.; Meyers, J.P. Three-dimensional modeling of electrochemical performance and heat generation of lithium-ion batteries
in tabbed planar configurations. J. Electrochem. Soc. 2011, 158, A835–A843. [CrossRef]
67. Zhang, J.; Yang, X.-G.; Sun, F.; Wang, Z.; Wang, C.-Y. An online heat generation estimation method for lithium-ion batteries using
dual-temperature measurements. Appl. Energy 2020, 272, 115262. [CrossRef]
68. Wang, P.; Yang, L.; Wang, H.; Tartakovsky, D.; Onori, S. Temperature estimation from current and voltage measurements in
lithium-ion battery systems. J. Energy Storage 2021, 34, 102133. [CrossRef]
69. Marelli, S.; Corno, M. Model-based estimation of lithium concentrations and temperature in batteries using soft-constrained dual
unscented kalman filtering. IEEE Trans. Control. Syst. Technol. 2021, 29, 926–933. [CrossRef]
70. Mahamud, R.; Park, C. Reciprocating air flow for li-ion battery thermal management to improve temperature uniformity.
71. Park, C.; Jaura, A.K. Dynamic Thermal Model of Li-Ion Battery for Predictive Behavior in Hybrid and Fuel Cell Vehicles; SAE Technical
Paper Series; SAE International: Warrendale, PA, USA, 2003.
72. Chen, L.; Hu, M.; Cao, K.; Li, S.; Su, Z.; Jin, G.; Fu, C. Core temperature estimation based on electro-thermal model of lithium-ion
batteries. Int. J. Energy Res. 2020, 44, 5320–5333. [CrossRef]
73. Maleki, H.; Shamsuri, A.K. Thermal analysis and modeling of a notebook computer battery. J. Power Sources 2003, 115, 131–136.
[CrossRef]
74. Lin, X.; Perez, H.E.; Mohan, S.; Siegel, J.; Stefanopoulou, A.G.; Ding, Y.; Castanier, M.P. A lumped-parameter electro-thermal
model for cylindrical batteries. J. Power Sources 2014, 257, 1–11. [CrossRef]
75. Li, D.; Yang, L. Identification of spatial temperature gradient in large format lithium battery using a multilayer thermal model.
Int. J. Energy Res. 2019, 44, 282–297. [CrossRef]
76. Lin, X.; Stefanopoulou, A.G.; Perez, H.E.; Siegel, J.B.; Li, Y.; Anderson, R.D. Quadruple adaptive observer of the core temperature
in cylindrical li-ion batteries and their health monitoring. In Proceedings of the 2012 American Control Conference (ACC),
Montreal, QC, Canada, 27–29 June 2012. [CrossRef]
77. Lin, X.; Fu, H.; Perez, H.E.; Siege, J.B.; Stefanopoulou, A.G.; Ding, Y.; Castanier, M.P. Parameterization and observability analysis
of scalable battery clusters for onboard thermal management. Oil Gas Sci. Technol. 2013, 68, 165–178. [CrossRef]
78. Sun, J.; Wei, G.; Pei, L.; Lu, R.; Song, K.; Wu, C.; Zhu, C. Online internal temperature estimation for lithium-ion batteries based on
kalman filter. Energies 2015, 8, 4400–4415. [CrossRef]
79
Energies 2021, 14, 5960
79. Dai, H.; Zhu, L.; Zhu, J.; Wei, X.; Sun, Z. Adaptive kalman filtering based internal temperature estimation with an equivalent
electrical network thermal model for hard-cased batteries. J. Power Sources 2015, 293, 351–365. [CrossRef]
80. Doughty, D.H.; Butler, P.C.; Jungst, R.G.; Roth, E. Lithium battery thermal models. J. Power Sources 2002, 110, 357–363. [CrossRef]
81. Xie, Y.; Li, W.; Hu, X.; Lin, X.; Zhang, Y.; Dan, D.; Feng, F.; Liu, B.; Li, K. An enhanced online temperature estimation for
lithium-ion batteries. IEEE Trans. Transp. Electrif. 2020, 6, 375–390. [CrossRef]
82. Sun, L.; Sun, W.; You, F. Core temperature modelling and monitoring of lithium-ion battery in the presence of sensor bias.
Appl. Energy 2020, 271, 115243. [CrossRef]
83. Zhu, C.; Shang, Y.; Lu, F.; Jiang, Y.; Cheng, C.; Mi, C. Core temperature estimation for self-heating automotive lithium-ion batteries
in cold climates. IEEE Trans. Ind. Inform. 2019, 16, 3366–3375. [CrossRef]
84. Xiao, Y. Model-based virtual thermal sensors for lithium-ion battery in EV applications. IEEE Trans. Ind. Electron. 2014, 62,
3112–3122. [CrossRef]
85. Ma, Y.; Cui, Y.; Mou, H.; Gao, J.; Chen, H. Core temperature estimation of lithium-ion battery for EVs using kalman filter.
Appl. Therm. Eng. 2020, 168, 114816. [CrossRef]
86. Ismail, N.H.F.; Toha, S.F.; Azubir, N.A.M.; Ishak, N.H.M.; Hassan, I.D.M.K.; Ibrahim, B.S.K. Simplified heat generation model for
lithium ion battery used in electric vehicle. IOP Conf. Ser. Mater. Sci. Eng. 2013, 53, 012014. [CrossRef]
87. Jeon, D.H. Numerical modeling of lithium ion battery for predicting thermal behavior in a cylindrical cell. Curr. Appl. Phys. 2014,
14, 196–205. [CrossRef]
88. Baba, N.; Yoshida, H.; Nagaoka, M.; Okuda, C.; Kawauchi, S. Numerical simulation of thermal behavior of lithium-ion secondary
batteries using the enhanced single particle model. J. Power Sources 2014, 252, 214–228. [CrossRef]
89. Du, S.; Jia, M.; Cheng, Y.; Tang, Y.; Zhang, H.; Ai, L.; Zhang, K.; Lai, Y. Study on the thermal behaviors of power lithium iron
phosphate (LFP) aluminum-laminated battery with different tab configurations. Int. J. Therm. Sci. 2015, 89, 327–336. [CrossRef]
90. Yi, J.; Lee, J.; Shin, C.B.; Han, T.; Park, S. Modeling of the transient behaviors of a lithium-ion battery during dynamic cycling.
91. Fleckenstein, M.; Bohlen, O.; Roscher, M.A.; Bäker, B. Current density and state of charge inhomogeneities in li-ion battery cells
with LiFePO4 as cathode material due to temperature gradients. J. Power Sources 2011, 196, 4769–4778. [CrossRef]
92. Santhanagopalan, S.; Ramadass, P.; Zhang, J.Z. Analysis of internal short-circuit in a lithium ion cell. J. Power Sources 2009, 194,
93. Li, Z.; Zhang, J.; Wu, B.; Huang, J.; Nie, Z.; Sun, Y.; An, F.; Wu, N. Examining temporal and spatial variations of internal
temperature in large-format laminated battery with embedded thermocouples. J. Power Sources 2013, 241, 536–553. [CrossRef]
94. Mutyala, M.S.K.; Zhao, J.; Li, J.; Pan, H.; Yuan, C.; Li, X. In Situ temperature measurement in lithium ion battery by transferable
flexible thin film thermocouples. J. Power Sources 2014, 260, 43–49. [CrossRef]
95. Srinivasan, R.; Carkhuff, B.G.; Butler, M.H.; Baisden, A.C. Instantaneous measurement of the internal temperature in lithium-ion
rechargeable cells. Electrochim. Acta 2011, 56, 6198–6204. [CrossRef]
96. Srinivasan, R. Monitoring dynamic thermal behavior of the carbon anode in a lithium-ion cell using a four-probe technique.
97. Schmidt, J.P.; Arnold, S.; Loges, A.; Werner, D.; Wetzel, T.; Ivers-Tiffée, E. Measurement of the internal cell temperature via.
impedance: Evaluation and application of a new method. J. Power Sources 2013, 243, 110–117. [CrossRef]
98. Troxler, Y.; Wu, B.; Marinescu, M.; Yufit, V.; Patel, Y.; Marquis, A.J.; Brandon, N.P.; Offer, G. The effect of thermal gradients on the
performance of lithium-ion batteries. J. Power Sources 2014, 247, 1018–1025. [CrossRef]
99. Brandon, N.; Mitcheson, P.; Yufit, V.; Howey, D.; Offer, G. Battery Monitoring in Electric Vehicles, Hybrid Electric Vehicles and
Other Applications. U.S. Patent 2,013,022,915,6A1, 8 May 2013.
100. Richardson, R.R.; Howey, D.A. Sensorless battery internal temperature estimation using a kalman filter with impedance
measurement. IEEE Trans. Sustain. Energy 2015, 6, 1190–1199. [CrossRef]
101. Zhu, J.; Sun, Z.; Wei, X.; Dai, H. A new lithium-ion battery internal temperature on-line estimate method based on electrochemical
impedance spectroscopy measurement. J. Power Sources 2015, 274, 990–1004. [CrossRef]
102. Debert, M.; Colin, G.; Bloch, G.; Chamaillard, Y. An observer looks at the cell temperature in automotive battery packs.
Control Eng. Pract. 2013, 21, 1035–1042. [CrossRef]
103. Hu, X.; Yuan, H.; Zou, C.; Li, Z.; Zhang, L. Co-estimation of state of charge and state of health for lithium-ion batteries based on
fractional-order calculus. IEEE Trans. Veh. Technol. 2018, 67, 10319–10329. [CrossRef]
104. Wang, Z.; Ma, J.; Zhang, L. State-of-health estimation for lithium-ion batteries based on the multi-island genetic algorithm and
the gaussian process regression. IEEE Access 2017, 5, 21286–21295. [CrossRef]
105. Hande, A. Internal battery temperature estimation using series battery resistance measurements during cold temperatures.
106. Howey, D.; Mitcheson, P.D.; Yufit, V.; Offer, G.; Brandon, N.P. Online measurement of battery impedance using motor controller
excitation. IEEE Trans. Veh. Technol. 2014, 63, 2557–2566. [CrossRef]
107. Liu, Z.; Li, H. A Spatiotemporal estimation method for temperature distribution in lithium-ion batteries. IEEE Trans. Ind. Inform.
2014, 10, 2300–2307. [CrossRef]
108. Khelif, R.; Chebel-Morello, B.; Malinowski, S.; Laajili, E.; Fnaiech, F.; Zerhouni, N. Direct remaining useful life estimation based
on support vector regression. IEEE Trans. Ind. Electron. 2017, 64, 2276–2285. [CrossRef]
80
Energies 2021, 14, 5960
109. Sbarufatti, C.; Corbetta, M.; Giglio, M.; Cadini, F. Adaptive prognosis of lithium-ion batteries based on the combination of particle
filters and radial basis function neural networks. J. Power Sources 2017, 344, 128–140. [CrossRef]
110. Liu, K.; Li, K.; Peng, Q.; Guo, Y.; Zhang, L. Data-driven hybrid internal temperature estimation approach for battery thermal
management. Complexity 2018, 2018, 1–15. [CrossRef]
111. Feng, F.; Teng, S.; Liu, K.; Xie, J.; Xie, Y.; Liu, B.; Li, K. Co-estimation of lithium-ion battery state of charge and state of temperature
based on a hybrid electrochemical-thermal-neural-network model. J. Power Sources 2020, 455, 227935. [CrossRef]
112. Li, Y.; Liu, W.; Zhong, J. Comparison of noninvasive and remote temperature estimation employing magnetic nanoparticles in DC
and AC applied fields. In Proceedings of the 2012 IEEE International Instrumentation and Measurement Technology Conference,
Graz, Austria, 13–16 May 2012.
113. Zhong, J.; Dieckhoff, J.; Schilling, M.; Ludwig, F. Influence of static magnetic field strength on the temperature resolution of a
magnetic nanoparticle thermometer. J. Appl. Phys. 2016, 120, 143902. [CrossRef]
114. Yoshida, T.; Enpuku, K. Simulation and quantitative clarification of ac susceptibility of magnetic fluid in nonlinear brownian
relaxation region. Jpn. J. Appl. Phys. 2009, 48, 127002. [CrossRef]
115. Zhong, J.; Liu, W.; Du, Z.; De Morais, P.C.; Xiang, Q.; Xie, Q. A noninvasive, remote and precise method for temperature and
concentration estimation using magnetic nanoparticles. Nanotechnology 2012, 23, 075703. [CrossRef]
116. Zou, D.; Li, M.; Wang, D.; Li, N.; Su, R.; Zhang, P.; Gan, Y.; Cheng, J. Temperature estimation of lithium-ion battery based on an
improved magnetic nanoparticle thermometer. IEEE Access 2020, 8, 135491–135498. [CrossRef]
81
energies
Article
All-SiC ANPC Submodule for an Advanced 1.5 kV EV
Charging System under Various Modulation Methods
Rafał Kopacz, Michał Harasimczuk, Bartosz Lasek, Rafał Miśkiewicz and Jacek Rabkowski
˛ *
Institute of Control and Industrial Electronics, Warsaw University of Technology, 00-662 Warsaw, Poland;
rafal.kopacz@pw.edu.pl (R.K.); michal.harasimczuk@pw.edu.pl (M.H.); bartosz.lasek@pw.edu.pl (B.L.);
rafal.miskiewicz@pw.edu.pl (R.M.)
* Correspondence: jacek.rabkowski@pw.edu.pl
Abstract: This work is focused on the design and experimental validation of the all-SiC active
neutral-point clamped (ANPC) submodule for an advanced electric vehicle (EV) charging station.
The topology of the station is based on a three-wire bipolar DC bus (±750 V) connecting an ac
grid converter, isolated DC-DC converters, and a non-isolated DC-DC converter with a battery
energy storage. Thus, in all types of power converters, the same three-level submodule may be
applied. In this paper, a submodule rated at 1/3 of the nominal power of the grid converter (20 kVA)
is discussed. In particular, four different modulation strategies for the 1.5 kV ANPC submodule,
exclusively employing fast silicon carbide (SiC) MOSFETs, are considered, and their impact on
the submodule performance is analyzed. Moreover, the simulation study is included. Finally, the
laboratory prototype is described and experimentally verified at a switching frequency of 64 kHz.
Citation: Kopacz, R.; Harasimczuk, It is shown that the system can operate with all of the modulations, while techniques PWM2 and
M.; Lasek, B.; Miśkiewicz, R.; PWM3 emerge as the most efficient, and alternating between them, depending on the load, should
Rabkowski,
˛ J. All-SiC ANPC be considered to maximize the efficiency. Furthermore, the results showcase that the impact of the
Submodule for an Advanced 1.5 kV different PWM techniques on switching oscillations, including overvoltages, can be nearly fully
EV Charging System under Various omitted for a parasitic inductance optimized circuit, and the choice of modulation should be based
Modulation Methods. Energies 2021, on power loss and/or other factors.
14, 5580. https://doi.org/10.3390/
en14175580 Keywords: ANPC converter; EV charging; multilevel converter; PWM methods; SiC MOSFETs
Academic Editors: Andrei Blinov,

Sheldon Williamson and
Hugo Morais
1. Introduction
Received: 11 July 2021 There is no doubt that easily available fast charging infrastructure is a necessary
Accepted: 30 August 2021 condition in the further expansion of electric vehicles (EVs) beyond current numbers, even
Published: 6 September 2021 in the most developed countries [1,2]. In comparison to traditional cars, the charging time
of EVs is, and will be, longer than refueling a tank with gasoline. However, fast charging
Publisher’s Note: MDPI stays neutral stations may offer a reduction in time from the range of hours to tens of minutes [3]. This
with regard to jurisdictional claims in is associated with an increase in charging power to hundreds of kWs, and, unfortunately, a
published maps and institutional affil- rising number of such stations is challenging for the power system. Therefore, an answer to
iations. this problem may be a battery energy storage, reducing power peaks during fast charging
periods [4,5]. Additionally, the storage may also act as local energy storage for a PV plant,
and perform short-time grid support services. All in all, the EV charging station with
energy storage becomes a high-power and complex power electronics system, as can be
Copyright: © 2021 by the authors. seen in Figure 1. Moreover, to decrease current levels and conduction losses in a common
Licensee MDPI, Basel, Switzerland. DC-link (750–800 V), a bipolar topology may also be taken into account [6]. All three
This article is an open access article types of power converters in such a system: grid-connected AC-DC, isolated DC-DC, and
distributed under the terms and non-isolated DC-DC, should be based on one of the multilevel topologies. Furthermore,
conditions of the Creative Commons in the optimal scenario, all of them should be based on the same topology to reduce the
Attribution (CC BY) license (https:// complexity and cost of the whole system, and to introduce modularity into the system.
creativecommons.org/licenses/by/ While the isolation stage of the AC-side is omitted in this paper, as the focus is on the
4.0/).

Energies 2021, 14, 5580
presented ANPC submodule, it is worth noting that either conventional low-frequency

transformers or solid-state transformers [7] are applicable in the EV charging system.
Figure 1. Bipolar DC grid-based EV charging station with energy storage.
Several multilevel converter topologies are applicable in EV charging systems [8].

Most notably, neutral point clamped (NPC), T-type, flying capacitor (FC), and modular
multilevel converter (MMC) topologies can be named. In the discussed case, the active
neutral point clamped (ANPC) topology of the submodule, introduced in [9], was selected
as it is characterized by more flexible control and the possibility to reach a more balanced
power loss distribution among the semiconductors compared to the conventional NPC
topology, while keeping the power device stress on a similar level [10,11]. Moreover,
when compared to other common multilevel structure, such as T-type and FC converter
topologies, the former is less efficient for the systems in such voltage range (1500 V) than
the NPC-derived topologies [12], whereas the latter cannot be employed in a system
with a bipolar DC bus for obvious reasons. Finally, when a comparison with MMC-
based systems [13] is considered, ANPC converters seem more appropriate, as MMC
introduces a bulkier structure, and adds further complexity to the system [8]. Furthermore,
ANPC systems were positively verified in terms of both three-wire [14] and EV charging
systems [15].
Moreover, as the SiC technology is constantly developing and providing power
semiconductor devices with superior performance compared to conventional Si coun-
terparts [16–18], especially for this specific voltage range, applying SiC MOSFETs to the
switches lead to, amongst many, increased efficiency and higher power density. However,
briefly after the commercial introduction of SiC power devices, when the SiC technology
was still relatively new, the cost of such power devices was significantly higher compared to
conventional Si IGBTs. Thus, different hybrid Si/SiC ANPC topologies were introduced, in
which SiC MOSFETs can be applied as the power devices for two or four out of six devices
per leg depending on the preferred PWM technique, showcasing satisfactory results in
terms of achieving a balance between cost and performance [18–22]. Nevertheless, even
though the hybrid topologies should be considered when the cost is taken into account,
the all-SiC system is still unmatched when strictly performance and maximization of
efficiency are considered [19]. Moreover, such a converter structure is and will become
even more compelling, as the wide-band-gap technology is currently becoming more and
more advanced, and thus SiC power devices will develop to be affordable to a greater
extent yearly.
Since the number of switches in the ANPC leg is high, so is the variety of PWM
methods applicable in such a system. Moreover, depending on the focus of a specific appli-
84
Energies 2021, 14, 5580
cation, there are modulation techniques that can target various factors, such as efficiency
or power density, ensuring equal loss distribution, or, finally lowering line filter require-
ments [10,23–25]. Furthermore, according to the literature, the impact of commutation path
lengths is a crucial matter, determining the proper PWM method for a specific application
as well. This is especially relevant when systems with SiC power devices are considered,
as wide-band-gap semiconductors are capable of high-speed switching, and thus are more
prone to ringing and overvoltages compared to its Si counterparts.
However, in this paper, except for validating the constructed low-volume prototype
of the all-SiC ANPC single leg rated at 1500 V DC and 6.67 kVA power (1/3 of three-phase
20 kVA system), it is shown that when enough care and focus is put into the design process
of the converter and thus the commutation path lengths are vastly minimized, the variances
between different modulation techniques in this regard are not as apparent and the choice
may be limited to other factors, namely in this case, efficiency. The conclusions are based on
a parasitic inductance optimized ANPC leg that can be used as a submodule to construct
full power electronic systems, e.g., three-phase bidirectional AC/DC converters as shown
in Figure 1.
Furthermore, in this paper, the PWM techniques are compared based on efficiency
and switching performance, strictly for an all-SiC system. In contrast, other researchers
have focused on a comparison between different Si/SiC configurations with strictly bound
modulation techniques, where each configuration was tested with its specific PWM method.
Finally, the conclusion is drawn that while all modulation techniques are viable, two emerge
as the most competent, one for lower power ratings and another for higher power ratings.
Thus, the assumption is made that to operate optimally, the modulation technique should be
changed according to the load. Moreover, while systems comprised of ANPC submodules
have been shown in the past, they are connected with other power semiconductor device
types, such as IGBTs [26,27] or IGCT [28]. There are no publications regarding SiC MOSFET-
based systems rated at MV level, whereas for such an application, the impact of parasitic
inductances due to high dv/dt rates and a high switching speed is much more severe, and
thus also more critical during the design process [29].
The paper is organized as follows. After the introduction, in Section 2, the basic
principles of the ANPC topology are explained together with the considered PWM methods
and their operation principles. Then, the simulation study is shown in Section 3, and, in
Section 4, the experimental model of the SiC-based submodule is presented along with
the results showcasing the experimental validation and further the discussion. Finally, the
paper is concluded with a summary in Section 5.
2. Modulation Strategies in Active Neutral Point Clamped (Anpc) Converter

The most popular inverter topology used in industrial power electronic is a basic
three-phase two-level (2-L) inverter [30]. This is mainly due to its simple design and
well-understood operation principles. However, the voltage stress of semiconductor power
devices in such topology is greater than in the DC-link voltage bus. This prevents the
use of 1.2 kV SiC power devices in 2-L inverters with greater DC voltage. A well-known
alternative to 2-L inverters are three-level (3-L) inverters [31]. The use of such a topology
provides a halved maximum voltage stress in semiconductor power devices. Furthermore,
due to the three-level nature of the output filter inductor voltage in 3-L inverters, it is
possible to reduce the harmonic distortion and output filter volume [32,33]. The low
switching time of SiC power MOSFETs also allows a reduction in switching losses. On the
other hand, the high value of stray inductance and di/dt of transistors during commutation
lead to voltage overshoots, which have a negative impact on MOSFETs’ lifetime, energy
conversion efficiency, and EMI. In 3-L inverters’ three different operation states (positive
state P, zero state 0, and negative state N) can be recognized, differing in voltage applied to
the inductor. A highly regarded 3-L inverter topology is the ANPC, used in the discussed
submodule. The ANPC converter consists of six active switches, S1 –S6 , connected according
to Figure 2.
85
Energies 2021, 14, 5580
Figure 2. A single-phase leg of the ANPC converter with SiC MOSFETs.
In the ANPC converter, when the output voltage is positive, the inverter is switching
between positive +V DC voltage and zero voltage; and when the output voltage is negative,
the inverter is switching between negative −V DC and zero voltage. Control of the converter
when the voltage is positive and negative is analogous. Therefore, in this article, the
different control methods are only described when the voltage is positive. P and N states
can be obtained only by turning on transistors S1 and S2 in P state, and S3 and S4 in N state.
During P state, transistor S6 can be turned on. Similarly, during N state, transistor S5 can
be on as well. This ensures constant vDS voltages equal to V DC on transistors S3 and S4
during P state, and on S1 and S2 during N state. Simultaneously, in the ANPC topology,
there are different approaches to obtain zero state. Four modulation strategies of the ANPC
and one of the NPC converters are depicted in Table 1 and Figure 3. In this paper, there
are four different modulations described (PWM1–PWM4). These PWM techniques differ
from each other in regard to the zero state, in which the current flows through different
conduction paths marked in Figure 4 by 2 and 3.
Table 1. Switching states of the ANPC inverter.
State S1 S2 S3 S4 S5 S6 Conduction Path(s) PWM Method

P 1 1 0 0 0 1/0 1 1, 2, 3, 4
0U3 1 0 1 0 0 1 3 2
0U2 0 1 0 0 1 0 2 1
0U1 0 1 0 1 1 0 2 4
0F 0 1 1 0 1 1 2 and 3 3
0L1 1 0 1 0 0 1 3 4
0L2 0 0 1 0 0 1 3 1
0L3 0 1 0 1 1 0 2 2
N 0 0 1 1 1/0 0 4 1, 2, 3, 4
86
Energies 2021, 14, 5580
(a) (b)
(c) (d)
Figure 3. Different modulation strategies for the ANPC converters (a) PWM 1, (b) PWM 2, (c) PWM
3, (d) PWM 4.
Figure 4. ANPC converter with depicted stray inductances and highlighted conduction paths.
In method PWM1 [34], during zero state transistors, S2 and S5 are on, and the current
flows through conduction path two (marked in blue in Figure 4). In this control method,
during transition P–0 a high value of di/dt in Lσ1 , Lσ5 , Lσ7 causes voltage spikes on the
switching transistors. In technique PWM2 [35], during zero state, the current flows via
transistors S3 and S6 , and in this case during the transition there is a high value of di/dt
in stray inductances Lσ1 –Lσ3 , and Lσ5 –Lσ7 . When we compare the transition in PWM2 to
87
Energies 2021, 14, 5580
the one in PWM1, the equivalent stray inductance is higher, and thus, di/dt is higher as
well, which leads to a higher value of voltage spikes. In method PWM3 [23,36], during
zero state, transistors S2 , S3 , S5 , and S6 are on, and current flows through conduction paths
two and three; equivalent resistances of both conduction paths are lower, which leads to
immensely lower conduction losses. PWM4 [22] is similar to PWM1. In both cases, after
the transition from P to 0, the current flows through conduction path two. However, in
PWM4 instead of transistor S6, transistor S4 is turned on, and thus, even if the modulation
pattern differs, the outcome is highly similar.
Generally, the described modulation strategies can be applied using SiC MOSFETs
and/or IGBTs in one inverter leg. For hybrid topologies mentioned in Section 1, MOSFETs
should be controlled with high frequency, while IGBTs should be switched with fundamen-
tal frequency to maximize the system performance. As mentioned before, SiC MOSFETs
are more expensive than IGBTs, and thus using different transistors in one module leads to
a reduced cost of the converter. However, as SiC MOSFETs become less and less expensive,
and the system exhibits better performance with all-SiC configuration. A system with
six SiC MOSFETs per leg is thus justified and interesting for further studies, especially
including the impact of various modulation techniques.
3. Simulation Study
The system in which the ANPC leg submodule was tested with the different PWM
techniques in this paper has been chosen as an open-loop single-phase inverter with a
resistive load, as it can mimic the converter’s behavior for a power factor near one quite
satisfactorily while keeping the circuitry simple. Thus, both simulation and experimental
tests were conducted in such a setup, according to system parameters shown in Table 2.
Table 2. System parameters.
Parameter Description
DC voltage 1500 V
AC voltage 230 V RMS/50 Hz
Rated power 6.67 kVA (1/3 of 20 kVA)
Operating frequency 64 kHz
SiC MOSFETs NTH4L040N120SC1
Filter inductor 220 μH
Filter capacitor 4.7 μF
DC capacitors 2 × 610 μF
At first, a simulation study in Plecs simulation software was performed to prelim-

inarily showcase the differences in the modulation methods, and establish the power
loss split between the converter components. The MOSFETs were modeled based on
datasheet values, including the impact of increased junction temperature and different
gate resistances. The other crucial source, namely the inductor, has also been included
in the power loss analysis, estimating the power loss as a sum of conduction power loss,
based on inductor resistance applied in the system based on a real model measurement.
Furthermore, component resistances, such as ESR, were added to ensure the converter’s
loss model is as accurate as possible. However, as Plecs does not simulate the transistor
switching processes fully, but rather operates on the basis of a lookup table with power
switching loss, the impact of parasitic inductances and ringing could not be observed in
the simulation study. Thus, the data from the simulation study were limited to power loss
determination and its split between the converter components.
The simulation tests were performed at near-nominal parameters (see Table 2) for
all five PWM methods described in Section 2. The impact of the different modulation
techniques on the total power loss and its distribution among the converter components
based on the simulation study can be observed in Figure 5. In general, for the system
parameters, PWM4 is the optimal technique for maximizing efficiency, as the total power
88
Energies 2021, 14, 5580
loss reached just 114 W, while other PWM methods (PWM1, PWM2, and PWM4) settled
close to each other at roughly 137 W.

P6
P6

P6
P6
P6

P6
P/

3:0 3:0 3:0 3:0
Figure 5. Power loss distribution among the converter components.
When we consider the power loss distribution among the converter components,
the situation differs quite notably for the semiconductor power devices, as shown in
Table 3, whereas the inductor exhibits nearly identical power loss for all the modulation
techniques. At first, PWM1 shows medium total power losses, in which MOSFETs S1 and
S4 are characterized by the highest value of roughly 30 W each, while pair S2 and S3 show
17.3 W, and the last pair S5 and S6 just 11.9 W per device. The second modulation method,
PWM2, is similar in terms of total power loss. However, it is also characterized by a highly
imbalanced distribution—transistors S2 and S3 are the sources of over a 70% semiconductor
power loss with 42.5 W, while pairs S1 , S4 , and S5 , S6 emit 6.4 and 10.8 W, respectively.
PWM3 exhibits top performance in terms of power loss, with nearly the most imbalanced
distribution, as over 60% of the power loss with 30.3 W is dissipated on pair S1 , S4 , while
pairs S2 , S3 , and S5 , S6 are the source of 11.8 and 5.6 W, respectively. However, it is worth
noting that the source of this higher imbalance compared with PWM1 and PWM4 is lower
power loss for the other MOSFET pairs, and not the increase in the S1 , S4 pair. Finally,
results for method PWM4 are very similar to PWM1 in terms of the loss distribution at 30.3,
16.4, and 11.7 W for transistor pairs S1 , S4 ; S2 , S3 ; and S5 , S6 , respectively. In terms of the
conducting paths and thus power losses, these methods are akin to each other.
Table 3. Power loss distribution among the ANPC submodule transistors.
V DC = 1500 V; P = 6.5 kW; vAC = 230 V rms

Parameter PWM1 PWM2 PWM3 PWM4
P(S1,S4) [W] 30.3 6.4 30.3 30.3
P(S2,S3) [W] 17.3 42.5 11.8 16.4
P(S5,S6) [W] 11.9 10.8 5.6 11.7
4. Experimental Study
4.1. The ANPC Submodule Prototype
In order to validate the system experimentally, the next step was to design and con-
struct the ANPC leg prototype. Since the system was to operate with a ±750 V DC bus, at
least 1200 V rated transistors were needed in the multilevel structure. Based on preliminary
calculations from the simulation study and analytical calculations, NTH4L040N120SC1 SiC
89
Energies 2021, 14, 5580
MOSFETs were chosen from a group of on the shelf, state-of-the-art power devices as all
of the switches, since these are characterized by satisfactory on-state resistance of 40 mΩ
and external Kelvin source connection. Therefore, they lead to minimized conduction and
switching losses and provide the possibility to switch the transistor in a fast and a robust
manner. Furthermore, as mentioned before, it was crucial to minimize the conduction loop
lengths in order to lower the effect of parasitic inductances that could lead to excessive
ringing, and cause overvoltages and increased power loss, which could result in working
outside the safe operating area of the power device and potentially even breakage. This was
achieved through the employment of a 4-layer power board structure along with a highly
compacted layout of the SiC MOSFETs, as well as additional 82 nF fast bypass capacitors,
put between +/0 and 0/− potentials as close to the power devices as possible. Since the
plan was to test several PWM methods, none of the conduction paths were favored, and
all were of similar length. However, such placement of the transistors leads to a situation
where the whole semiconductor power loss has to be dissipated in the near vicinity of the
center of the heatsink, thus leading to a less balanced heat distribution; in summary, more
capable power loss dissipation measures had to be used. Therefore, as the submodule was
to operate with a power of near 6.7 kW in a low-volume system, heatsink Fischer LAM 6
with a highly efficient 48 V fan was employed in the prototype shown in Figure 6.
Figure 6. ANPC submodule—photo of the prototype.
Moreover, the prototype consists of self-made gate drivers based on the UCC21750
chip from Texas Instruments, providing satisfactory switching performance as well as
fault protection measures. Finally, the component count of the experimental model of
the submodule concludes with two main DC capacitors rated at 800 V and 60 μF. The
constructed ANPC leg prototype is shown in Figure 6.
4.2. Experimental Setup

Alongside the ANPC submodule prototype, the experimental setup consisted of an
LC line filter, constructed from a 220 μH inductor and a 4.7 μF capacitor, as well as a
reconfigurable resistive load. Furthermore, since single-phase systems require high DC
capacitance, two additional 550 μF/900 V capacitors were added to support the built-in
submodule capacitances. The converter was controlled at 64 kHz operating frequency via
a DSP-based circuit, established on TMS320F28379D launchpad. The system was supplied
through a 2 kV/5 A DC power supply from Magna Power, and a Yokogawa WT5000
power analyzer was used to measure the efficiency. Finally, the waveforms were obtained
using Tektronix MSO56 oscilloscope with isolated voltage probes (Tektronix THDP0100
and P2505A) and a current probe (Tektronix TCP0030A). The scheme and the photo of the
experimental setup are shown in Figure 7.
90
Energies 2021, 14, 5580
$
V'& v'6
S
C+
S S
'& L i$&
3RZHU 9 $
VXSSO\ V'& S S
v
C 9 R v$&
C/ S
SKDVH$13&LQYHUWHU

I,1 I287
V,1 3RZHUDQDO\]HU V287

(a)
(b)
Figure 7. Experimental setup for the ANPC single-phase inverter system with a resistive load—(a) scheme,
(b) photo.
4.3. Results from the Experimental Study

The core focus of the performed tests was to validate the constructed ANPC sub-
module up to its nominal parameters, and establish the most proficient PWM technique.
Since in the whole EV system and the 1500 V DC bipolar bus is required to be connected
with the European grid (230 V RMS/50 Hz) at a rated power of 1/3 out of 20 kVA, the
modulation index m applied was equal to 0.45, resulting in a voltage gain of roughly 0.153.
Figure 8 presents an exemplary oscillogram with line frequency-focused view for PWM3
at 6.7 kW, showcasing ANPC leg output voltage v0 , DC-link voltages V DC1 and V DC2 ,
load AC voltage vAC , and current iAC . As the impact of the PWM method is minimal in
a 50 Hz context, it is assumed that for other modulation techniques, the waveforms are
identical and thus are not shown. As can be seen, there is still some imbalance between the
+/0 and 0/− DC voltages, regardless of high 610 μF capacitance; however, its impact is
limited regarding the load AC current and voltages as its THD settled below 5% for all the
tests, and thus we can omit the mismatches throughout the further result analysis. For the
nominal parameters, the AC load current settled at roughly 28 A, whereas the load voltage
was established close to 230 V RMS resulting in a power of 6.5 kW. This operating point
was further used as a nominal for further experimental comparison between the different
PWM methods.
91
Energies 2021, 14, 5580
Figure 8. Experimental results from a test at 1500 V DC and 6.5 kW power (m = 0.45, vAC = 230 V)
with a line frequency-focused view. From the top: ANPC leg output voltage v0 , DC-link voltages
V DC1 and V DC2 , load AC voltage vAC , and current iAC .
4.3.1. Transistor Overvoltage and Ringing

At first, a study on transistor overvoltage and ringing for different PWM methods
was conducted. Since the switching behavior of MOSFETs within the transistor pairs S1
and S4 , S2 and S3 , as well as S5 and S6 were identical, only the switching voltages for the
bottom switches (S3 , S4 , S6 ) were measured. Unfortunately, due to the highly compacted
design of the submodule, it was impossible to apply current probes and measure the
transistor currents. Nevertheless, in terms of the safety of operation for the semiconductor
power devices, the drain–source voltage is the crucial factor, while the impact of the
current oscillations was indirectly included in the study through efficiency measurements.
Furthermore, it is worth noting that transistor overvoltages were also affected by the
DC-link voltage imbalances. Thus, peak overshoot voltages could have been even more
limited with a higher capacitance, and would not occur if the submodule was used in a
different system, e.g., three-phase inverter, where DC-link voltage balancing is assured.
Figures 9 and 10 depict the experimental switching waveforms near the peak line
current for the system, operating at nominal values for the modulation technique with
the highest voltage oscillations (PWM2). As shown, even though the oscillation is visible,
it does not exceed 900 V, which is a safe value for the MOSFETs applied in the system.
Furthermore, it is worth noting that due to DC-link voltage imbalance, the waveforms for
the positive line current (Figure 9) are different from those obtained for the negative current
(Figure 10). This variance between DC-link voltages V DC1 and V DC2 settled at roughly 60 V,
corresponding to roughly 4% of nominal voltage.
(a)
Figure 9. Cont.
92
Energies 2021, 14, 5580
(b)
Figure 9. Exemplary experimental waveforms showcasing drain–source transistors for positive load
current (iAC > 0) for the modulation technique with highest overvoltages (PWM2)—(a) turn-on,
(b) turn-off.
(a)
(b)
Figure 10. Exemplary experimental waveforms showcasing drain–source transistors for negative
load current (iAC < 0) for the modulation technique with highest overvoltages (PWM2)—(a) turn-on,
(b) turn-off.
Peak overvoltage values for all the studied modulation techniques are showcased in
Table 4. In the previously shown figures, only switching near the peak value of the line
current was considered, whereas the data shown in Table 4 consist of the highest value
throughout the entire 50 Hz period. Thus, this data are the basis for considerations for all
the switches. Based on this data, we can observe that the peak overvoltage value difference
between the PWM methods reached maximally 8% of the nominal drain–source voltage of
750 V for PWM2, while the variances between PWM1, PWM3, and PWM4 were as low as
4% of the nominal voltage. The difference between PWM1 and PWM4 is the most visible
when transistor voltages are compared. For PWM1, transistors S1 and S4 (depending on
the line current sign) are not bound to any constant potential, but rather float depending on
the current ANPC leg state. This is not an issue for PWM4, as transistors S5 and S6 connect
the floating potential to the zero voltage, and thus peak transistor voltage overshoots are
93
Energies 2021, 14, 5580
lower. The significant difference between PWM2 and other techniques is caused by this
method’s relatively lengthy conduction loop, as mentioned in Section 2. Nevertheless,
the variance is still on a minimal level. Therefore, when a similar power rating as in the
presented system is considered, it is safe to assume that for a well-optimized system, in
terms of conduction path length, the effect of chosen PWM technique on the transistor
overvoltage is somewhat limited and should not be as important as other factors, such
as power loss and its distribution or filter requirements, or even omitted at all. However,
this effect is enlarged when the current is higher. Thus, such an approach should not be
applicable in very high power systems.
Table 4. Results for the different modulation patterns at nominal ratings of the ANPC leg.
V DC = 1500 V; P = 6.5 kW; vAC = 230 V rms

Parameter PWM1 PWM2 PWM3 PWM4
vDS_max(S3) [V] 832 957 846 827
vDS_max(S4) [V] 889 802 868 873
vDS_max(S6) [V] 857 848 842 843
PLOSS(exp.) [W] 182 177 165 182
PLOSS(sim.) [W] 137 137 114 136
4.3.2. Power Losses

The other crucial factor in which the PWM methods were compared is the system
efficiency. Since power loss could not be measured individually on every converter compo-
nent without impacting the inverter performance, total power loss was measured as the
difference between input and output converter power, according to Figure 7.
Figure 11a showcases the total inverter efficiency at nominal system voltages (1500 V
DC and 230 V AC), with a constant modulation index at 0.45 and varying load resistance,
so that the power could be measured in 20–100% range of its nominal value. As shown,
modulation techniques PWM3 and PWM2, similarly as in the simulation study, exhibit
the highest efficiency. For the nominal power, and thus with the highest transistor current,
PWM4 shows the lowest power losses, implying that the conduction losses are the main
source of power loss. This is since this modulation method employs all four middle
switches (S2 , S3 , S5 , S6 ), and thus, the lowest effective on resistance. PWM2, on the other
hand, showcases top performance for lower power, below 60% of the nominal value. In this
modulation type, when only the positive half of the line current is considered, transistors S1
and S5 do not switch at all as the switching occurs between S2 and S3 . Therefore, switching
loss is limited to this transistor pair, contrary to other PWM techniques where the switching
occurs for more power devices. However, when higher power is regarded, the importance
of switching loss is diminished, and the efficiency becomes very similar for all the methods
except for PWM3, which is characterized by more available conduction paths. Thus, using
PWM2 and PWM3 alternately, depending on the load conditions, should be considered
to ensure the lowest power losses throughout the whole operating range. Furthermore,
developing an algorithm that optimally chooses the modulation technique according to the
operating point may be considered. When we study the two other methods more generally,
in terms of efficiency, they are pretty similar with the slight advantage for PWM1 for lower
power ratings.
94
Energies 2021, 14, 5580

P/266 >:@
>@

3:0
3:0
3:0
3:0

3:0 3:0 3:0 3:0
P >:@
(a) (b)
Figure 11. Experimental characteristics showcasing performance of the ANPC leg in function of
converter power P—(a) efficiency at 1500 V DC, 230 V AC, and m = 0.45, (b) power loss at nominal
operating point (1500 V DC, 6.5 kW, m = 0.45).
Furthermore, to exhibit the difference in actual loss values rather than efficiency,
Figure 11b focuses on the power loss difference for different modulation techniques at
the nominal operating point with the full power of 6.5 kW. When we compare the results
from the experimental tests with the simulation study (see Table 4) we can observe that
the presumptions noted in Section 3 are confirmed via the experiments on the prototype
as well, with lowest power losses for PWM3 with 165 W. In contrast, techniques PWM1,
PWM2, and PWM4 settled close to each other at 182, 177, and 182 W respectively.
5. Conclusions
This paper presents an MV ANPC submodule with state-of-the-art SiC MOSFETs for
an advanced EV charging system. The constructed low-volume prototype of the all-SiC
ANPC leg rated at 1500 V DC and 6.67 kW power has been experimentally validated to
work with satisfactory switching performance and efficiency above 97.5% for the nominal
operating point, which is a substantial value for such a low modulation index and voltage
gain (1500 V DC to 230 V AC). Furthermore, as the design process’s focus was to minimize
the parasitic inductances in the converter, the transistor voltages and ringing were rela-
tively low below 5% of the steady-state value, so that satisfactory switching performance
could be achieved. Finally, several PWM techniques have been analyzed, tested, and
compared for the specific application shown in this paper, focusing on the impact of all the
presented modulation methods strictly for an all-SiC ANPC leg. This is in contrary to other
publications in the area, in which the different PWM techniques were applied, but only
in various SiC/Si hybrid ANPC leg configurations, usually limited to 1–2 modulations
per configuration.
The obtained results show that for an MV all-SiC ANPC inverter submodule rated
at 1.5 kV DC and applied in an advanced EV charging system, shown in Figure 1, PWM2
is the best for lower power, while PWM3 is the choice for a higher power (over 60% of
nominal value). However, for other power electronics applications, depending on the
required voltage levels (and thus the modulation index), as well as for other state-of-the-art
SiC MOSFETs and/or Si IGBTs, the outcomes may vary, as so would the ratio between
switching and conduction loss, power loss distribution among the components, as well as
other factors. Thus, efficiency wise it is not easy to strictly determine which PWM technique
is the optimal choice universally, as there are nuances for each application that can affect
the power loss quite noticeably. Furthermore, changing the modulation technique during
operation, depending on the load parameters should also be considered to achieve the
95
Energies 2021, 14, 5580
best performance, and thus, the highest efficiency. Nevertheless, based on the theoretical
and experimental performed studies for a parasitic inductance optimized system with
similar power ratings as in the presented ANPC leg, the impact of conduction loops on
transistor overvoltages and ringing between the different PWM methods, even for a quite
high voltage of 1.5 kV, is not crucial, and can be nearly fully omitted. Thus, the optimal
choice for the modulation technique should be limited to other required parameters, such
as efficiency.
Author Contributions: Conceptualization, R.K., M.H. and J.R.; methodology, M.H. and R.K.; simula-
tion study, R.K.; experimental validation, R.K., B.L. and M.H.; original draft preparation, R.K. and
M.H.; writing—review and editing, R.K., M.H., R.M. and J.R.; supervision, J.R.; project administra-
tion, J.R.; funding acquisition, J.R. All authors have read and agreed to the published version of the
manuscript.
Funding: The research leading to these results has received funding from the EEA/Norway Grants
2014–2021.
References
1. Rivera, S.; Kouro, S.; Vazquez, S.; Goetz, S.M.; Lizana, R.; Romero-Cadaval, E. Electric Vehicle Charging Infrastructure: From Grid
to Battery. IEEE Ind. Electron. Mag. 2021, 15, 37–51. [CrossRef]
2. Ghosh, A. Possibilities and Challenges for the Inclusion of the Electric Vehicle (EV) to Reduce the Carbon Footprint in the
Transport Sector: A Review. Energies 2020, 13, 2602. [CrossRef]
3. Ronanki, D.; Kelkar, A.; Williamson, S.S. Extreme Fast Charging Technology—Prospects to Enhance Sustainable Electric Trans-
portation. Energies 2019, 12, 3721. [CrossRef]
4. Hussain, A.; Bui, V.-H.; Baek, J.-W.; Kim, H.-M. Stationary Energy Storage System for Fast EV Charging Stations: Simultaneous
Sizing of Battery and Converter. Energies 2019, 12, 4516. [CrossRef]
5. Rafi, M.A.H.; Bauman, J. A Comprehensive Review of DC Fast-Charging Stations With Energy Storage: Architectures, Power
Converters, and Analysis. IEEE Trans. Transp. Electrif. 2021, 7, 345–368. [CrossRef]
6. Rivera, S.; Lizana, R.; Kouro, S.; Dragičević, T.; Wu, B. Bipolar DC Power Conversion: State-of-the-Art and Emerging Technologies.
IEEE J. Emerg. Sel. Top. Power Electron. 2021, 9, 1192–1204. [CrossRef]
7. Tahir, Y.; Khan, I.; Rahman, S.; Nadeem, M.F.; Iqbal, A.; Xu, Y.; Rafi, M. A state-of-the-art review on topologies and control
techniques of solid-state transformers for electric vehicle extreme fast charging. IET Power Electron. 2021, 14, 1560–1576. [CrossRef]
8. Poorfakhraei, A.; Narimani, M.; Emadi, A. A Review of Multilevel Inverter Topologies in Electric Vehicles: Current Status and
Future Trends. IEEE Open J. Power Electron. 2021, 2, 155–170. [CrossRef]
9. Bruckner, T.; Bemet, S. Loss balancing in three-level voltage source inverters applying active NPC switches. In Proceedings of the
2001 IEEE 32nd Annual Power Electronics Specialists Conference (IEEE Cat. No.01CH37230), Vancouver, BC, Canada, 17–21 June
2001; Volume 1132, pp. 1135–1140.
10. Ma, L.; Kerekes, T.; Rodriguez, P.; Jin, X.; Teodorescu, R.; Liserre, M. A New PWM Strategy for Grid-Connected Half-Bridge Active
NPC Converters With Losses Distribution Balancing Mechanism. IEEE Trans. Power Electron. 2015, 30, 5331–5340. [CrossRef]
11. Bruckner, T.; Bernet, S.; Guldner, H. The active NPC converter and its loss-balancing control. IEEE Trans. Ind. Electron. 2005, 52,
12. Schweizer, M.; Kolar, J.W. Design and Implementation of a Highly Efficient Three-Level T-Type Converter for Low-Voltage
Applications. IEEE Trans. Power Electron. 2013, 28, 899–907. [CrossRef]
13. Ronanki, D.; Williamson, S.S. Modular Multilevel Converters for Transportation Electrification: Challenges and Opportunities.
IEEE Trans. Transp. Electrif. 2018, 4, 399–407. [CrossRef]
14. Teston, S.A.; Vilerá, K.V.; Mezaroba, M.; Rech, C. Control System Development for the Three-Ports ANPC Converter. Energies
2020, 13, 3967. [CrossRef]
15. Lara, J.; Masisi, L.; Hernandez, C.; Arjona, M.A.; Chandra, A. Novel Five-Level ANPC Bidirectional Converter for Power Quality
Enhancement during G2V/V2G Operation of Cascaded EV Charger. Energies 2021, 14, 2650. [CrossRef]
16. Rabkowski, J.; Peftitsis, D.; Nee, H. Silicon carbide power transistors: A new era in power electronics is initiated. IEEE Ind.
Electron. Mag. 2012, 6, 17–26. [CrossRef]
17. Biela, J.; Schweizer, M.; Waffler, S.; Kolar, J.W. Sic versus si—Evaluation of potentials for performance improvement of inverter
and dc–dc converter systems by sic power semiconductors. IEEE Trans. Ind. Electron. 2011, 58, 2872–2882. [CrossRef]
96
Energies 2021, 14, 5580
18. Feng, Z.; Zhang, X.; Yu, S.; Zhuang, J. Comparative Study of 2SiC&4Si Hybrid Configuration Schemes in ANPC Inverter. IEEE
Access 2020, 8, 33934–33943. [CrossRef]
19. Zhang, L.; Lou, X.; Li, C.; Wu, F.; Gu, Y.; Chen, G.; Xu, D. Evaluation of Different Si/SiC Hybrid Three-Level Active NPC Inverters
for High Power Density. IEEE Trans. Power Electron. 2020, 35, 8224–8236. [CrossRef]
20. Feng, Z.; Zhang, X.; Wang, J.; Yu, S. A High-Efficiency Three-Level ANPC Inverter Based on Hybrid SiC and Si Devices. Energies
2020, 13, 1159. [CrossRef]
21. Zhang, L.; Liu, S.; Chen, G.; Yang, X. Evaluation of Hybrid Si/SiC Three-Level Active Neutral-Point-Clamped Inverters. In
Proceedings of the 2019 IEEE 28th International Symposium on Industrial Electronics (ISIE), Vancouver, BC, Canada, 12–14 June
2019; pp. 840–845. [CrossRef]
22. Zhang, D.; He, J.; Pan, D. A Megawatt-Scale Medium-Voltage High Efficiency High Power Density “SiC+Si” Hybrid Three-Level
ANPC Inverter for Aircraft Hybrid-Electric Propulsion Systems. In Proceedings of the 2018 IEEE Energy Conversion Congress
and Exposition (ECCE), Portland, OR, USA, 23–27 September 2018; pp. 806–813.
23. Jiao, Y.; Lee, F.C. New Modulation Scheme for Three-Level Active Neutral-Point-Clamped Converter with Loss and Stress
Reduction. IEEE Trans. Ind. Electron. 2015, 62, 5468–5479. [CrossRef]
24. Floricau, D.; Gateau, G.; Leredde, A.; Teodorescu, R. The efficiency of three-level Active NPC converter for different PWM
strategies. In Proceedings of the 2009 13th European Conference on Power Electronics and Applications, Barcelona, Spain, 8–10
September 2009; pp. 1–9.
25. Belkhode, S.; Shukla, A.; Doolla, S. Enhanced Hybrid Active-Neutral-Point-Clamped Converter With Optimized Loss Distribution-
Based Modulation Scheme. IEEE Trans. Power Electron. 2021, 36, 3600–3612. [CrossRef]
26. Mayor, A.; Rizo, M.; Monter, A.R.; Bueno, E.J. Commutation Behavior Analysis of a Dual 3L-ANPC-VSC Phase-Leg PEBB Using
4.5-kV and 1.5-kA HV-IGBT Modules. IEEE Trans. Power Electron. 2019, 34, 1125–1141. [CrossRef]
27. Jiao, Y.; Lu, S.; Lee, F.C. Switching Performance Optimization of a High Power High Frequency Three-Level Active Neutral Point
Clamped Phase Leg. IEEE Trans. Power Electron. 2014, 29, 3255–3266. [CrossRef]
28. Apeldoorn, O.; Odegard, B.; Steimer, P.; Bernet, S. A 16 MVA ANPC-PEBB with 6 kA IGCTs. In Proceedings of the Fourtieth IAS
Annual Meeting. Conference Record of the 2005 Industry Applications Conference, Kowloon, Hong Kong, China, 2–6 October
2005; Volume 812, pp. 818–824.
29. Chen, M.; Pan, D.; Wang, H.; Wang, X.; Blaabjerg, F. Investigation of Switching Oscillations for Silicon Carbide MOSFETs in
Three-Level Active Neutral-Point-Clamped Inverters. IEEE J. Emerg. Sel. Top. Power Electron. 2021, 9, 4839–4853. [CrossRef]
30. Jahns, T.M.; Dai, H. The past, present, and future of power electronics integration technology in motor drives. CPSS Trans. Power
Electron. Appl. 2017, 2, 197–216. [CrossRef]
31. Cittanti, D.; Guacci, M.; Mirić, S.; Bojoi, R.; Kolar, J.W. Comparative Evaluation of 800V DC-Link Three-Phase Two/Three-Level
SiC Inverter Concepts for Next-Generation Variable Speed Drives. In Proceedings of the 2020 23rd International Conference on
Electrical Machines and Systems (ICEMS), Hamamatsu, Japan, 24–27 November 2020; pp. 1699–1704.
32. Teichmann, R.; Bernet, S. A comparison of three-level converters versus two-level converters for low-voltage drives, traction, and
utility applications. IEEE Trans. Ind. Appl. 2005, 41, 855–865. [CrossRef]
33. Schweizer, M.; Friedli, T.; Kolar, J.W. Comparative Evaluation of Advanced Three-Phase Three-Level Inverter/Converter
Topologies Against Two-Level Systems. IEEE Trans. Ind. Electron. 2013, 60, 5515–5527. [CrossRef]
34. He, J.; Zhang, D.; Pan, D. An Improved PWM Strategy for “SiC+Si” Three-Level Active Neutral Point Clamped Converter in
High-Power High-Frequency Applications. In Proceedings of the 2018 IEEE Energy Conversion Congress and Exposition (ECCE),
Portland, OR, USA, 23–27 September 2018; pp. 5235–5241.
35. Guan, Q.; Li, C.; Zhang, Y.; Wang, S.; Xu, D.D.; Li, W.; Ma, H. An Extremely High Efficient Three-Level Active Neutral-Point-
Clamped Converter Comprising SiC and Si Hybrid Power Stages. IEEE Trans. Power Electron. 2018, 33, 8341–8352. [CrossRef]
36. Barater, D.; Concari, C.; Buticchi, G.; Gurpinar, E.; De, D.; Castellazzi, A. Performance Evaluation of a Three-Level ANPC
Photovoltaic Grid-Connected Inverter With 650-V SiC Devices and Optimized PWM. IEEE Trans. Ind. Appl. 2016, 52, 2475–2485.
[CrossRef]
97
energies
Article
Interface Converters for Residential Battery Energy Storage
Systems: Practices, Difficulties and Prospects
Ilya A. Galkin 1, *, Andrei Blinov 2 , Maxim Vorobyov 1 , Alexander Bubovich 1 , Rodions Saltanovs 1
and Dimosthenis Peftitsis 3
1 Faculty of Electrical and Environmental Engineering, Riga Technical University, LV1048 Riga, Latvia;
maksims.vorobjovs@rtu.lv (M.V.); aleksandrs.bubovics@rtu.lv (A.B.); rodions.saltanovs@rtu.lv (R.S.)
2 Department of Electrical Power Engineering and Mechatronics, Tallinn University of Technology,
19086 Tallinn, Estonia; andrei.blinov@taltech.ee
3 Department of Electrical Power Engineering, Norwegian University of Science and Technology,
NO-7491 Trondheim, Norway; dimosthenis.peftitsis@ntnu.no
* Correspondence: gia@eef.rtu.lv
Abstract: Recent trends in building energy systems such as local renewable energy generation have
created a distinct demand for energy storage systems to reduce the influence and dependency on
the electric power grid. Under the current market conditions, a range of commercially available
residential energy storage systems with batteries has been produced. This paper addresses the
area of energy storage systems from multiple directions to provide a broader view on the state-
of-the-art developments and trends in the field. Present standards and associated limitations of
storage implementation are briefly described, followed by the analysis of parameters and features of
commercial battery systems for residential applications. Further, the power electronic converters are
reviewed in detail, with the focus on existing and perspective non-isolated solutions. The analysis
Citation: Galkin, I.A.; Blinov, A.;
covers well-known standard topologies, including buck-boost and bridge, as well as emerging
Vorobyov, M.; Bubovich, A.;
Saltanovs, R.; Peftitsis, D. Interface
solutions based on the unfolding inverter and fractional/partial power converters. Finally, trends
Converters for Residential Battery and future prospects of the residential battery storage technologies are evaluated.
Energy Storage Systems: Practices,
Difficulties and Prospects. Energies Keywords: residential energy storage; battery energy storage systems; standards; grid interface
2021, 14, 3365. https://doi.org/ converters; intellectual property; bidirectional converters; AC-DC power converters; DC-DC power
10.3390/en14123365 converters; multilevel converters; partial power converters
Academic Editor: Teuvo Suntio
Received: 31 March 2021 1. Introduction

Accepted: 2 June 2021
Consumption of resources as well as their collection and processing are usually
Published: 8 June 2021
uneven. First of the all, it involves energy resources, traditionally, food and various fissile
fuels. Nowadays, the necessity to store energy has gained new forms that are applied to
the energy resources, specific for the dedicated technology equipment. This, in particular,
regards electrical engineering, the rapid development of which during the last two centuries
iations.
has formed the demand for storages of electrical energy even at the level of residential
applications. During recent years, this tendency has become more topical due to several
reasons. Firstly, renewable energy sources are in much wider use. In addition, this use is
obliged by some administrative regulations like EU directives [1–3]. In spite of the irregular
generation profile, the renewable energy sources are being installed even at the households.
Secondly, the range and number of various household devices have expanded. There exist
plenty of storages dedicated to electrical energy [4]. For example, it is possible to convert
electrical energy into chemical (in the form of pure hydrogen) by means of electrolysis and
then back—by means of a fuel cell [5]. However, in spite of the most recent achievements
in the field of fuel cells [6,7] and development of converter technologies for fuel cells [8],
creativecommons.org/licenses/by/ the most functional, reliable and energy efficient equipment for electrical energy is an
4.0/). electrochemical battery energy storage (BES) system.

Energies 2021, 14, 3365
The constantly increasing number of papers (Figure 1) devoted to battery energy

storage systems (BESSs) proves the importance of these energy storage devices in various
applications. These papers address all aspects of their use, but particular attention is
paid to the interface converters of BESSs. The numerous review papers devoted to this
topic [9–12] describe a generalized state of the art in this field. Typically, they evaluate
which converter schemes are more energy efficient, with a reduced component count and
lower voltage/current stresses. At the same time, the role and peculiarities of the interface
converters in the context of the BESS structure are usually not clear-cut and detailed in
these reports.
150
120
90
60
30
0
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
Conferences Journals
Figure 1. Number of recent IEEE publications about BESS.
BESSs nowadays are also readily commercially available. The analysis of the market
of household electrical equipment [13,14] shows that numerous BESSs are already available
as a market offering. On the one hand, the variety of their parameters and operation
conditions provides wide choices; on the other hand, it makes the choice more complicated
for the final users of BESS and complicates the development of the interface converters
for different BESSs. In addition, the elaboration and commercialization of BESSs and their
interface converters have a strong link to the market of some renewable energy sources
and pure electric vehicles, which may not only act as BESSs, but also, after their recycling,
provide high voltage (HV) second-life Li-ion batteries for use in BESS [15,16].
The goal of this work is to analyze the majority of interface converters in the context
of the corresponding BESSs, their operation conditions (standards, energy tariffs, subsidies
and other elements of energy policy), BESS market trends and after this analysis, to formu-
late prospective development directions of the BESS interface converters. In particular, this
regards the converter schemes for HV batteries.
The rest of the paper is organized in five sections. Section 2 reviews the motivating
factors of the BESS study: battery technologies, their applications, as well as standards and
other regulations that may regard this work. Section 3 briefly analyzes the commercially
available BESSs, trying to emphasize their internal structure. Section 4 provides a broad
analysis of converter technologies applicable to BESSs. Section 5 discusses the previously
analyzed equipment and technologies in the context of BESS development. Finally, the
conclusions are given in Section 6.
2. Motivation and Driving Factors for Use of Battery Energy Storage Systems
2.1. Development of Electrochemical Energy Storages
The most intensive development of electrochemical batteries has taken place since
the late 20th century and it is still progressing. Due to the constantly growing demand for
portable electronics, vehicular technologies and energy systems, the battery technologies of
100
Energies 2021, 14, 3365
known electrochemistry have been “polished” and new technologies have been introduced
to the market. Presently, the most significant commercially available battery technologies
are [17,18]: advanced lead-acid (LA), nickel-oxyhydroxide (NiMH), sodium–sulfur (NaS),
various kinds of Li-ion batteries, as well as redox flow batteries (RFBs), in particular,
vanadium redox batteries (VRBs) [18]. LA technology, the oldest among them, is still the
cheapest as well as quite energy efficient (up to 85%). The drawbacks of LA batteries are
rather low specific energy (Figure 1) and low number of charge-discharge cycles (lifetime).
Historically, the next successive NiMH technology (replacement for NiCd) is characterized
by average specific power, specific energy and lifetime, but undergoes significant self-
discharge and is of low charge-discharge efficiency (65%). The NaS batteries are of high
specific energy, energy efficiency and lifetime (90% and 4000 cycles, respectively [18]), but
their operation temperature is high—they require heating, which makes them impractical
in many cases. Today, the most quickly developing battery technology is the Li-Ion. Its
high specific energy, specific power (Figure 2), lifetime (up to 10k cycles), energy efficiency
(up to 95%) achieved at reasonable price makes the technology very suitable for use in
portable electronics, all-electric vehicles, household energy systems, and, even, in energy
distribution grids [19]. However, the specific parameters of Li-Ion batteries depend on
relevant chemistry and all advantages are typically not concentrated in one device. Finally,
RFBs, in particular VRBs, are the batteries that utilize reduction–oxidation reaction between
two liquids, which occurs through a membrane. The liquids are pumped to the membrane
that makes RFBs similar to fuel cells, where the liquids are chemically restorable. The
main advantage of these batteries is their potentially infinite lifetime. Lastly, it must be
mentioned that modern batteries are not just a series connection of galvanic cells. They
often include electronics for balancing, management and protection as well as chargers in
some cases. Therefore, these batteries can be considered as complex complete energy units
for immediate use [20–22].
250
Li-ion
LiNiCoAlO2
200
NaS
Specific Energy [Wh/kg]
LiNiMnCoO2
150
L
LiFe2PO4
100
NiMH
(NiCd)
50
LA
500 1000 1500 2000

Specific Power [W/kg]
Figure 2. Specific energy and specific power of the commercially available batteries (based on
data from [18]).
2.2. Extensive Use of Battery Energy Storages in Transport

One of the recent global societal and legislative tendencies on the national and in-
ternational levels is the request to reduce the consumption of fossil fuels and to increase
the efficiency of energy consumption [3,23]. Among other areas, this involves vehicular
technologies as well. Regarding ground vehicles, this initiative means wider use of plug-in
electric vehicles (PEVs) or all-electric vehicles (AEVs) and hybrid ones, equally in the
public and private sectors. In [24,25], the availability of cost-effective batteries of several
101
Energies 2021, 14, 3365
hundred volts for main electrochemical energy storage of PEV is reported. More recent
papers [26–29] consider these PEV storage systems valuable enough to be a part of the
energy supply grid. Further development of the BESs makes their use possible in larger
ground vehicles—first of all, in the public transport [30,31].
Better BESs are also required for water vehicles, first of all, for smaller auxiliary
vehicles—boats, yachts, water buses, etc. For example, in [32], the electrification of the
water buses in Venice is considered as a successful example of BES use in water transport.
At the same time, with regard to bigger ships and vessels, the role of BESs differs with time.
While earlier configurations of marine energy systems utilize high voltage batteries for
stabilizing the traditional on-board AC grid and power smoothing [33,34], modern systems
also take into account the possibility of all-electric propulsion of the ship [34,35].
Finally, the most advanced BESs are applicable in aircraft. The traditional electrical
supply of an aircraft combines an AC and DC grid. Better performance of the applied
batteries leads to a better quality of the 28 V DC grid [36,37]. At the same time, top BES tech-
nologies allow production of extremely light batteries that enable all-electric aircraft [38].
In conclusion, the extensive use of batteries in transport, in particular, the growing
number of light PEVs, high capacity of their batteries and huge capacity of these batteries
in total, as well as their wide distribution, make these BESs a substantial grid resource
for storing energy. These BESs and their interface converters are typically high-voltage
devices, but the corresponding solutions of the interface converters can be adopted for
residential use.
2.3. Recent Challenges in the Field of Power and Energy Supply

The request to reduce fossil fuel consumption [3,23] regards also power distribution
and supply networks. For the power and energy supply systems, this means that the
burning of fossil fuels must be substituted with renewable energy generation. In turn, the
main properties of renewable energy generation are:
(1) Uneven generation profile—regardless of the kind, the renewable energy sources
typically do not provide constant power. In particular, the generation of PV panels
depends on solar irradiation and varies with the daytime, cloudiness, season, location
of PV and solar activity. The generation of wind turbines depends on the wind
strength, which is unique for its location, season and occasional weather fluctuation.
The generation of hydro and waves turbines depends on the amount of water that is
a long-term function of seasonal and global weather changes.
(2) Variety of power ratings and types of energy sources exist even within the same
group. For example, the power of PV depends on the local properties and financial
abilities of a particular household.
(3) Variety of allocation of the renewable energy sources—depending on the particular eco-
nomic conditions and policy of energy operator, these sources may be allocated differently.
Altogether, this makes renewable energy generation less stable and reliable. This, as
well as several other problems [9,39–43], can be solved with the help of Battery Energy
Storage Systems (BESSs).Figure 3 shows the use of BESSs in energy applications.
When considering a BESS in a small household with different loads and renewable
energy sources, it is very important to smoothen renewable energy generation—providing
storage for excessive renewable or cheap grid energy [44–46]. The BESS is also capable of
performing the function of an uninterruptible power supply. This is the main function in
the case of islanded residential grids [47–49]. At very uneven loads, the BESS may also
smoothen the real-time loading of supply equipment—transformers and lines.
In the distribution grids, the functions of the dedicated BESS are similar but more
specified. Price compensation now can be considered as a complete function of energy
trading, smoothening of power generation regards not only renewables, but smoothening
of consumed power at this level saves the capacity of distribution equipment. Additionally,
BESS in distribution grids may perform grid service functions: grid black restart as well
as voltage and frequency regulation [39,50]. The choice of BESS parameters is a subject
102
Energies 2021, 14, 3365
of multiple factors [51]: standards, power losses, voltage of majority of available PEVs,
compatibility with pure resistive loads.
BESSs in Power and Energy Supply
BESSs in Supply
(Local, Household, Residential)
l)
l)
BESSs in
Grids
Plug-in
Connected BESSs in Distribution Grid
Electric
to Distribution Islanded Vehicles
Network
Problems to solve: Problems to solve:

Smoothing of Renewable Generation Energy Trading
Smoothing of Energy Price Voltage Regulation
Supply Continuity Frequency Regulation
Smoothing of Consumed Power Smoothing Power Generation
Grid Forming (in Islanded Grid) Equipment Capacity
Power Quality (harmonics, reactive power) Harmonic Compensation
Black Start
Figure 3. Battery Energy Storage Systems (BESSs) in power and energy supply at a glance.
2.4. Standards and Other Regulations Applicable to Battery Energy Storage Systems
The standards directly related to the electrical energy storage systems of households
are still under development. In Europe, this is being done by the IEC 120 committee
group [52]. They have developed a roadmap for developing standards, which is planned
to be completed by the end of 2023. Until that date, European manufacturers have to use
general standards for the production of power converters, in particular, power interfaces
for alternative energy sources and uninterruptible power supply (Table 1).
Table 1. Summary of Standards and Regulations applicable to BESS.
Reference Application Area of the Standard

[53] USA, Converter housing and selection of components
[54] IEC, Classification of BESS locations in households
[55] IEC, Voltage inverters for high voltage DC networks
[56] IEC, Controlling of converters in microgrids
IEC, Connection of PV to the grid and requirements for electromagnetic
[57]
compatibility parameters
IEC, Bidirectional low voltage (up to 1000 V AC and 1500 V DC) converters
[58]
connected to the grid and description of the terms used in these networks
IEC, Test methods and acceptable parameters for low voltage uninterruptible
[59,60]
power supplies
[61] IEC, Disposal of converters of uninterruptible power supplies
[62] USA, Safety regulations within data centers and telecom central offices
In the USA, a universal standard has been developed that describes the operation of
electrical energy converters in distributed networks. With regard to BESS, the manufac-
turers also have to apply general standards for converters. This includes standards for
interface converters of energy storage. In addition, in the USA, the parameters of batteries
are defined and standardized and based on the standards of telecommunication equipment
(Table 1).
103
Energies 2021, 14, 3365
3. Commercially Available Residential Storage Systems

In this section, the BESs available on the market are analyzed taking into account the
parameters available from the product datasheets or application manuals. Despite the
market for such devices still being dynamic, some common properties and features can
already be distinguished as common practice in the field.
3.1. Typical Example of Battery Energy Storage Systems Dedicated to Household Applications
The Tesla Powerwall 1 (3.3 kW/6.4 kWh) was one of the first attempts to include
BESS into a household energy system and has been available on the market since 2015.
It operates with a DC-bus and, in general, has to be installed in conjunction with a grid
inverter, which is sold separately.
This precluded its use as a completely independent BESS, reduced market prospects
and shortly led to its replacement by the Tesla Powerwall 2 (5 kW/13.2 kWh) [63]. In
contrast to the previous model, the Powerwall 2 (Figure 4a) includes an AC inverter and
can be connected directly to the AC grid. This enables its use as a residential BESSs,
regardless of the renewable generation source (solar panels or a wind generator). Therefore,
the functional features of Powerwall 2 have expanded significantly, including the possibility
of stand-alone operation without grid connection (islanded mode). For normal operation,
it requires an additional commutation unit called “energy gateway” and its full cycle
efficiency is 90%.
(a) (b)
(c) (d)
Figure 4. Typical examples of system configurations of different BESs: (a) Tesla Powerwall 2, (b) SonnenBatterie, (c) Adara
Power-Residential and (d) Sunverge energy.
Sonnen is another early market player that began offering its residential BESSs in
December 2015 [64]. These BESSs are designed for households with solar and wind power
generators providing energy storage and backup power. They are available in two versions,
with a built-in inverter for PVs (hybrid output) and without it (eco output). In Germany,
the company launched a coordination network that brings together power producers
and storage owners. This service allows the participants resided in the same network to
exchange electricity with each other, exporting surplus to the grid. Currently, this service
104
Energies 2021, 14, 3365
has over 10,000 users. With LiFePO4 batteries in its system, the manufacturer claims an
output power of 2.5–3.3 kW in the “eco output” version and 5.5 kW in the “hybrid output”
version. The energy capacity of the base model is 5 kWh with the ability to increase it up
to 15 kWh in 2.5 kWh steps. The manufacturer promises a 98% maximum efficiency of
the batteries and a 96% efficiency of the converter, which gives a total cycle efficiency of
around 88.5%. The internal topology of SonnenBatterie and SonnenFlat is not disclosed,
but the structural diagram of their operational environment (Figure 4b) shows that it is
connected to the main supply grid as well as to the secondary grid formed by the solar
panel inverter through an automatic transfer switch (ATS). This enables a SonnenBatterie
to operate in an uninterruptible power supply mode.
Enphase Energy is another company that entered the residential storage market with
its “AC Battery” in 2015. It is a very compact (0.27 kW/1.2 kWh) modular system that can
be used in conjunction with micro-inverters and the “Envoy-S gateway” [65]. Later, the
company’s storage portfolio was extended with the Encharge 3 (1.28 kW/3.3 kWh) [66]
and Encharge 10 [67], which is composed of three of the former units. According to the
datasheet information, the cycle efficiency of a newer Enphase product is 89% at half
power. Backup power from the battery can be provided using an additional microgrid
interconnection device.
Other notable market players are Victron Energy with a range of products like Easy
Solar and MultiPlus [68]; Adara Power’s Residential [69,70] coupled with an inverter from
Schneider Electric (Figure 4c) [71] and Sunverge Energy (Figure 4d) [72]. Moreover, one
of the key market players is the battery manufacturer LG Chem [73], who is offering its
low- and high-voltage battery modules for integration with SMA, Fronius, SolarEdge, and
Huawei inverters/chargers.
3.2. Summary of Parameters and Features of Commercial Residential BESs

Due to the market dynamics, with both large and small companies are entering and
leaving the market continuously, so it is hard to determine a global leader in the area.
Moreover, some of the products currently have a limited proposal or are available only in
certain regions. The typical price for typical residential BESs is currently in the range of
1–2 kEUR/kWh (Table 2). Technical information on these products is mostly limited—only
general specifications are typically available. Still, certain common properties of residential
BESSs can already be distinguished. In the majority of cases, the utilized energy storage
is a low-voltage (50 V) Li-ion battery, which is associated with relatively high currents.
Although the particular topology configurations used in these systems are not revealed by
the manufacturers, such voltage level would in general require a rather complex interface
converter featuring a transformer for the required voltage step-up. Using RESU10 and
RESU10H from LG as a reference, one of the reasons for using a battery with such voltage
level is its 14% reduced price, as compared with the higher voltage battery of the same
energy capacity. This results in round-trip efficiencies of most residential BESS being
around 90%, which seems to be a current technological limit for such configurations.
The current market of BESSs shows a clear trend of their transformation from the
auxiliary BESSs, complementing a solar or wind farm with a smoothing energy storage
(AC and DC coupling), towards a complete energy system with BES that does not depend
on the availability of alternative energy (only AC coupling). While the BESSs of the first
type have a DC output and often need a separate grid converter, newer BESSs of the second
type are intended for AC operation due to the intrinsic AC interface. From the point of
view of their features and functions, the earlier BESSs were focused on local power supply
and equalization or shift of peak consumption, but the newer systems have a range of
advanced functions, for example, integration on energy system level, i.e., possibility of
combining several household grids with BESS into a distributed power plant.
105
Energies 2021, 14, 3365
Table 2. Summary of BESS for general use and use with renewable energy sources available on the market.
BESS Manufacturer/Model Maximal Energy Charge/Discharge

Capacity [kWh] BES Power [kW] Battery Voltage [V] Coupling Reference
Tesla PowerWall 13.5 5 50 AC [63]

Sonnen Batterie Eco 15 3.3 48 AC [64]
Adara Power (Residential) 20 12 50 AC/DC [70]
Sunverge Modular up to 19.4 6 48 AC/DC [72]
Solax X-ESS G4 or Hybrid X1/X3 Stackable up to 23
4 300 AC/DC [74]
+ Triple Power (BES) (4 modules)
SolarEdge + RESU10H 9.8 5 400 AC/DC [75]
PowerVault 3 20 3.3/5.5 52 AC [76]
Puredrive Storage II AC 5 kWh 5/10 3 50 AC [77]
Duracell Energy Bank 3.3 3.3 52 AC [78]
Enphase Encharge 3 3.5 1.3
67 AC [79,80]
Enphase Encharge 10 10.5 3.8
Nissan/Eaton xStorage 4.2 . . . 10 3.6 . . . 6 90 AC/DC [81]
Samsung SDI All in One 3.6 4.6 60 AC/DC [82]
Varta Pulse/Pulse Neo 3 3.3 1.6/1.4 50 AC [83]
Varta Pulse/Pulse Neo 6 6.5 2.5/2.3
Sunny Boy Storage External battery 3.7/5/6 360 AC [84]
Victron Energy EasySolar External battery 0.9/1.7/3.5 12.8–51.2 DC [85]
3.3. Isolated Converters of Commercially Available Residential BESSs

As it was shown in Sections 3.1 and 3.2, most of the commercially available BESSs
utilize a low voltage battery (see Table 2 for details). The use of such a low-voltage battery
while maintaining, at the same time, good control performance, requires that the entire
BES interface converter or part of it be a low voltage circuit that, in turn, typically means
the use of an isolation transformer. The use of the transformer also allows satisfying
the potential safety requirements (see Section 2.4 for details). The transformer may be a
network transformer operating at the frequency of the supply grid or a high-frequency
transformer. Both solutions have benefits and disadvantages briefly considered below.
3.3.1. Converters with Grid-Frequency Isolating Transformers at AC Side

In general, adding of a transformer at the grid side moves the entire semiconductor
circuitry into a low voltage operation, but its topology may be almost of any type, as
presented below in Section 4. Therefore, there are two large groups of converters with a
network transformer: single stage converters (Figure 5a) and converters with two conver-
sion stages (Figure 5b). The BESS may also be equipped with a transformer at the request
of the operator and/or legal regulations, in order to meet the operational requirements.
This, however, regards more to BESSs for distribution grids, in particular, ABB with the
ESSPro product line [86] and NIDEC with the Silcolstart product line [87]. The transformer
installed at the AC side makes the operation of the converter possible at lower voltages,
but makes the BESS heavier and bulky.
LV/HV HV Grid LV/HV

BES Converter BES Converter HV Grid
er
Transformer Transformer
er
DC DC Passive DC Passive
DC AC filter AC filter
(a) (b)
Figure 5. BESS power electronics converters with a transformer: (a) single-stage, (b) two-stage.
3.3.2. Converters with High-Frequency Isolating Transformers

A high-frequency transformer may be allocated in the DC-link (Figure 6a). The most
versatile and straightforward kind of the implementation of this approach is the use the
circuitry known as dual active bridge (DAB, Figure 6b). Classical DAB [88] is a hard-
switching topology that, compared with non-isolated interface converters, considered
106
Energies 2021, 14, 3365
below in Section 4, is less reliable and energy efficient due to the extra components as well
as due to its hard-switching nature. However, if combined with a soft-switching technique,
for example, applying a resonant network, it may operate with better efficiency [89,90].
Grid inverter + LV Vbatt HV Vbus + Grid Frontend

Isolation HF
DC DC
AC
LV Battery
Grid
HF LF
LV BES
AC DC AC
DC-DC converter DC Link – –
(a) (b)
Figure 6. High-frequency transformers in the DC-link of the interface converter of LV BES: (a) functional diagram,
(b) transformer in conjunction with classical DAB.
In a more advanced approach (Figure 7a), the high-frequency transformer is located at

the edge of the DC-link and the AC grid [91]. This requires that the AC part of the topology
contains bidirectional switches so that it can operate at both polarities of the grid voltage.
The performance of this topology can be improved with the help of resonant networks (red
elements in Figure 7b) and advanced modulation methods [92]. Similar topologies and
their properties are well described in [93].
+ LV Vbatt
LV BES
Isolation
HF
DC AC
LV Battery
Grid
AC Grid
HF LF
AC AC
HFLC
–
(a) (b)
Figure 7. Full-bridge DC-AC converter with a high-frequency transformer: (a) functional diagram, (b) converter derived
from DAB.
The abovementioned DC-DC and DC-AC converters contain an energy-bypassing

transformer. Alternatively, the high frequency link may contain also a storage transformer
(split coil). In the most explicit form, this storage transformer is seen in a flyback converter.
This converter, however, is a DC-DC circuit and its use, therefore, is directly possible only in
the DC-link [94] similar to DAB (Figure 6b). At the same time, adopting of the principle to
AC networks is possible. For example, the converter presented in [95] contains two flyback
converters dedicated to positive and negative half-waves. The inputs of the converters
are connected in parallel to a low voltage battery, but their outputs—in series to the grid
(Figure 8). The interface converters with a high-frequency storage transformer have the
same drawbacks as original flyback converters: rather low power and highly pulsating
current on both sides (including battery side) that requires sufficient filtering.
107
Energies 2021, 14, 3365
LV BES AC grid
+
Filter
Figure 8. Inverter with a high-frequency transformer derived from a flyback converter.
4. Topologies of Non-Isolated Interface Converters for High-Voltage Battery Energy

Storage Systems
One of the ways to overcome some limitations of the existing residential BESS is to
utilize a battery with higher voltage (~200–500 V) and enable the use of a simpler and more
efficient interface converter. In fact, some companies, like SolaX, SMA and SunnyBoy, are
already on this path. Due to massive electrification of transportation where higher voltage
batteries are used to reduce charging current and time, the cost for higher voltage batteries
should decline further and make the use of high voltage (HV) batteries more feasible for
residential BESSs.
This section is devoted to the analysis of existing and perspective non-isolated power
electronic interfaces that can be applicable to the residential HV BESs. The main goal is to
highlight the benefits and limitations of various configurations and assess their feasibility
and performance. In addition to the standard single-, two-stage and multilevel topologies,
emerging configurations like impedance-source, partial and fractional power converters
are analyzed.
4.1. Functions and Structure of Interface Converters for BES

According to the analysis of commercially available residential BESSs, two main
configurations can be distinguished: DC- and AC-coupled. The first group is generally
represented by the power electronic systems that are often referred to as “hybrid inverters”
(Figure 9a). They allow integration of both PV and battery into a single multiport unit.
Such solutions are well-suited for new installations, but the choice of suitable storage
configurations could be limited. On the other hand, the AC-coupled storages are often
stand-alone systems that are directly connected to the residential AC grid (Figure 9b).
In general, such solutions are more flexible, as they can be integrated into any existing
installation. However, for such systems, charging of a battery from a PV typically involves
more energy conversion stages, with a negative impact on efficiency.
The interface converter of a BES needs to perform two main functions, along with a
range of auxiliary application-based functions. The main functions of the BES are sinusoidal
shaping of the AC grid current and forming the DC current of the BES in both directions
of power flow. The abovementioned functions can be implemented in a single stage
bidirectional DC-AC inverter/rectifier; however, such solutions are typically overall less
efficient due to battery voltage variation as compared to two-stage systems [96]. Therefore,
the BES interface is usually comprised of a bidirectional DC-DC stage that is interfaced
with a battery, followed by the DC-AC inverter/rectifier. The state-of-the-art and other
potential configurations of power electronic interfaces for HV BES are analyzed in the
following sections.
4.2. Single Stage DC-AC Bidirectional Inverters/Rectifiers

This section presents the state-of-the-art and emerging single-stage grid-tie
inverter/rectifier topologies. The main goal of these systems is to convert a DC voltage
into the sinusoidal AC waveform and vice versa. Most commercial systems require the
108
Energies 2021, 14, 3365
DC voltage to be relatively stable, with their value higher than the amplitude of the grid
voltage, while some of the emerging topologies potentially offer enhanced flexibility.

(a) (b)
Figure 9. Coupling of units in residential energy systems: (a) DC coupling, (b) AC coupling.
4.2.1. Bridge Converters

A high voltage battery can be attached to the grid through a single stage bidirectional
(four-quadrant) interface converter. The most frequently mentioned converter is a transistor
bridge. Such bridge itself is a parallel connection of two (three) transistor legs with two
transistors (and anti-parallel diodes) in each (Figure 10). The converter includes also an
inductance coil between its AC terminals and the grid implementing an AC current source
(Figure 10a). A diagonal couple of transistors and the couple of diodes located in the
opposite diagonal form a chopper capable of converting the grid voltage at its particular
polarity. In Figure 10a, red elements represent the chopper for the positive half-wave,
but blue elements—for the negative. The chopper is bidirectional and can be considered
as a buck converter supplied from the DC bus or as a boost converter supplied from
the AC grid [97]. One transistor leg can be substituted by a series connected capacitors
(capacitor leg), thus forming a transistor-capacitor bridge (Figure 10b), more frequently
named “half-bridge” [97]. In this topology, it is also possible to identify two choppers for
both half-waves of AC voltage. Finally, it is possible to apply this approach of schematic
synthesis to three phase systems (Figure 10c,d). This forms the three-phase transistor
bridge and the three-phase transistor bridge with a capacitor leg coupled to the grid via
inductor-based AC current source [98].
4.2.2. Topologies without Explicit Bridge

The intrinsic choppers shown in Figure 10 can be deployed without forming an explicit
bridge. This scheme is defined as a dual-buck grid converter known since 1997 [99,100].
With this approach, the elements of the “positive” and “negative” choppers are different,
which enables them to be further optimized.
Figure 11 shows how the elements of implicit choppers are extracted (red elements—
for positive and blue elements—for negative). With this approach, the inductance coil is
not shared between “positive” and “negative” branches (Figure 11a). These coils can be
magnetically coupled (Figure 11b), providing their lower weight/volume and therefore,
lower weight/volume of the converter itself [101], without the reduction of the performance
and reliability of the converter. Alternatively, both branches can be combined through a
couple of series connected diodes (Figure 11c) [102], keeping the same advantages.
109
Energies 2021, 14, 3365
HV BES
HV BES
+ – + –
– + – +
Grid Grid
(a) (b)
Grid
Grid
HV BES
HV BES
(c) (d)
Figure 10. Configurations of single-stage bridge rectifiers-inverters for BESs: (a) AC current sourced transistor bridge,
(b) AC current sourced transistor/capacitor bridge (half-bridge), (c,d) three-phase schemes.
Grid
Grid
Grid
HV BES
HV BES
HV BES
(a) (b) (c)
Figure 11. Dual-buck grid converter: (a) derived from half-bridge, (b) magnetically coupled branches, (c) diode coupled.
Extracting of the intrinsic voltage converters at both grid terminals of a full-bridge

converter forms another kind of the dual-buck grid converter (Figure 12a) [102]. Another
version derived from the full-bridge topology can be synthesized by means of direct
combining of two DC sourced buck converters—attached to each terminal of the grid [76].
In this converter, the switches located at the grid side are continuously conducting at
the corresponding grid voltage polarity that reduces switching losses. Additionally, such
converter may be “tied to positive voltage node” (as shown in Figure 12b) or “ground
tied”. Finally, adding two diodes at the grid side (Figure 12c) allows operating in “ground
tied” and “positive node tied” modes [103,104], making the operation of the switches
more symmetrical. The converter shown in Figure 12b can be equipped with magnetically
coupled inductance coils or coupling diodes, as shown in Figure 11.
110
Energies 2021, 14, 3365
Grid
Grid
Grid
HV BES
HV BES
HV BES
(a) (b) (c)
Figure 12. Dual-buck grid converters: (a) derived from full-bridge, (b,c) synthesized of two combined buck converters.
Due to a low number of elements, the considered single-stage converters operate with
best efficiency at their particular operation point. However, when considered in conjunction
with the attached battery, their efficiency is not outstanding and drops significantly at other
operation points due to the higher losses in the converter and the battery [86]. In addition,
these converters operate as an AC grid supplied boost or a BES supplied buck converter
that requires minimal battery voltage to be higher than the amplitude of the grid voltage.
4.2.3. Multilevel Converters

Multilevel converters (MLC) can be considered as a specific kind of the single-stage
converters, processing energy in separate cells of a BESS battery. In contrast to the above-
considered topologies that always deal with the same DC voltage or with the entire battery,
the multilevel converters form their output of DC voltage that may have several levels
obtained directly from the battery. The advantages of multilevel converters are lower
harmonic distortion, switching losses and electromagnetic interference [105]. There are
three main topologies of multilevel converters: cascaded H-bridge converters (also known
as multilevel converters with independent sources), neutral point clamped multilevel
converters (also known as diode clamped multilevel converters), and multilevel converters
with flying capacitors.
Cascaded H-Bridge Converter Structures

In the case of cascaded H-bridge multilevel converters, each phase contains several
series-connected modules (Figure 13a) composed of dedicated cells and an inverter, to-
gether forming an independent source. Within a BESS [106], these sources can be charged
and discharged more evenly due to the independent nature of their involvement in the
current path and potentially free exchange of the sources [107].
There exist various types of power converters and energy storage building blocks.
The most common converter is a single-phase transistor bridge (H-bridges) shown in
Figure 14a, which generates AC voltage on its output, thus controlling the charge or
discharge process of the connected battery cells. Another typical configuration given in
Figure 14b includes an AC generating H-bridge in conjunction with a synchronous buck
converter that compensates voltage changes in the cell(s).
Lastly, a successful commercial implementation of BESS with a multilevel converter
was offered by SolarEdge [108]. It is based on a multilevel DC converter with multiple
DC modules connected in series (Figure 13b). Allocation of the multilevel structure in the
DC bus enables significant simplification of the cell converters (Figure 14c). The DC/DC
converters can operate in the following modes: balancing circuit, charger and battery
discharger. In turn, if the DC bus is formed by an MLC, the grid frontend can be a simple
commutation matrix or an efficient pulse mode inverter or a short-circuit proof converter
with an impedance network. A similar topology developed by ABB for distribution
networks [109] includes an array of complex cells containing two transistors and a battery
111
Energies 2021, 14, 3365
with switches and capacitors. A cell may work as a boost or buck converter and is capable
of shunting the cell if needed.
Grid
Energy Storage Energy Storage Energy Storage
and Inverter and Inverter and Inverter
Building Block Building Block Building Block
(a)
Grid
Grid Inverter
Energy Storage Energy Storage Energy Storage

+ DC Converter + DC Converter + DC Converter
(b)
Figure 13. Functional diagrams of cascaded H-Bridge multilevel converters: (a) traditional configura-
tion of AC MLC, (b) configuration DC MLC with unipolar cell converters and grid frontend.
Vbatt0
to BESS to BESS
Vbatt0
Vbatt0
to
phase phase
BESS
phase
(a) (b) (c)
Figure 14. Power converters for multilevel converter building blocks: (a) full bridge or H-bridge, (b) H-bridge with
correcting synchronous buck converter, (c) unipolar bidirectional converter [108].
Neutral Point Clamped Multilevel Converters

The simplest kind of the neutral point clamped multilevel converters is known as
the diode clamped topology (Figure 15a). It has quite high efficiency compared to other
topologies. However, there are some disadvantages: the number of power diodes is
quadratic related to the level count, which makes this topology quite difficult to use when
a large number of levels is needed. Another disadvantage of the topology is that charge
balancing in the capacitors is needed. Another type of the neutral point clamped multilevel
converter is an active clamped multilevel converter shown in Figure 15b. Additional
switches enable the distribution of power losses more evenly between the switches. Besides,
it is possible to synthesize 0 V level differently, providing different charge/discharge paths.
112
Energies 2021, 14, 3365
Grid Grid Grid
HV BES
HV BES
HV BES
(a) (b) (c)
Figure 15. Multilevel converters with solid HV battery: (a) diode clamped, (b) active clamped, (c) flying capacitor.
The use of neutral point clamped converters in the BESS system is described in [110,111].
Reference [112] demonstrates the use of neutral point clamped and active neutral point
clamped converters in BESSs. In [113], an overview of modular converters (including
active neutral point clamped converters) in BESS systems is given. Diode clamped and
independent source multilevel converters in BESS applications, indicating also larger
operating range of the diode clamped converters, are compared in [114].
Multilevel Inverter with Flying Capacitors

The main difference between multilevel converters with neutral point clamped and
multilevel converters with flying capacitors (Figure 15c) is that instead of clamping diodes,
capacitors are used. Similar to the diode-clamped topology, the main disadvantage of the
multilevel inverter with flying capacitors is the large number of used capacitors, which
makes the practical implementation of this solution larger in terms of packaging. In spite
of this drawback, some recent papers report that the topology itself can be successfully
applied in BESS based on GaN switches: [115] presents a BESS with a 13-level converter,
but [116]—a 9-level converter for aircraft. In addition, the BESS interface converter offered
by SolarEdge in [108] utilizes the MLC with flying capacitors as a grid inverter.
4.3. Impedance-Source Bidirectional Inverters/Rectifiers

The problems of conventional topologies related to the battery voltage variation can be
mitigated with the family of impedance source (IS) converters. These topologies incorporate
a special network, which allows step-up of the input voltage using a shoot-through state in
the inverter bridge, which is a prohibited condition in conventional inverters. As a result,
IS converters can be less prone to short-circuit faults. There is a variety of impedance source
networks proposed in the literature for a range of applications with different properties
and features (Figure 16), including three-phase and multilevel configurations [117]. The
majority of basic impedance source topologies were initially unidirectional; however, some
studies address the bidirectional versions potentially suitable for residential BES [118,119].
4.4. Bidirectional Two-Stage DC-AC Converters

This section presents the state-of-the-art and emerging power electronic interfaces for
BES, featuring two explicit stages. In a general case, the first stage is a bidirectional DC-DC
converter, which processes varying battery voltage and controls the charge/discharge
current. It operates in conjunction with the DC-AC inverter/rectifier addressed in the
previous section, which provides interface with the grid.
113
Energies 2021, 14, 3365
Grid
BES
1f/3f full-bridge
Z-source 1f/3f half-bridge
Push-pull + unfolder
qZ-source
Synchronous buck + unfolder
Matrix converter
etc.
Y-source Trans-Z-source
Impedance Source Bidirectional Commutation Array
Figure 16. Generalized configuration of the BES interface converter with an impedance source.
In the most obvious operation mode, this DC-DC converter provides the stabilized
voltage in the DC-link at all operation points of the battery (Figure 17a) while the rectifier-
inverter modulates the voltage at the grid end according to the phase of the network
voltage and required grid current. It was demonstrated that two-stage configurations are
overall superior to the single DC-AC inverter/rectifier in terms of efficiency throughout
the battery voltage range [86]. Moreover, the stable DC-link voltage allows integration
of other DC sources and loads; therefore, such solutions can be suited for both DC- and
AC-coupled BESS.
VDC(t)
VGR(t)
VGR(t)
VDC(t)
VGR(t)
VDC(t)
VBT(t)
t t t
t t t t
Rectifier- Buck- Polarity Half-wave
Inverter boost converter shaper
Grid BES Grid BES
DC DC DC DC
Filter
AC DC AC DC
(a) (b)
Figure 17. Operation and configuration of BES interface converters with two stages: (a) common DC-link, (b) unfolding topology.
However, one more operation mode and configuration is possible. In this mode,
the DC-DC converter forms unipolar sine half-waves in the DC-link, but the rectifier-
inverter applies the formed half-waves to the grid with the correct polarity (Figure 17b).
In the second case, the rectifier-inverter does not operate in a real switch-mode—it just
commutates the half-waves at the grid frequency. Therefore, in this operation mode, the
switching losses of the rectifier-inverter are negligible, while the grid filter can be omitted
or reduced due to the continuous profile of the voltage at the grid port of the rectifier-
inverter [120]. In the single-phase configuration, the AC-DC converter is a bridge or
half-bridge circuit, close to that shown in Figure 16 without the inductance coil. Depending
on the required power and input connection, it can be a single-phase [120–122] or a
three-phase [122,123] circuit.
4.4.1. Two-Stage Converters with Stabilized DC-Link

The typical configuration of a two-stage converter assumes voltage stabilization at
an intermediate DC-link to compensate battery voltage variation and provide optimal
operating conditions for the DC-AC inverter/rectifier. Such configuration can be suited for
both DC- and AC-coupled BES. The standard DC-DC stage topologies include buck, boost,
buck-boost, etc. The common disadvantage of these standard configurations is that both
stages have to be rated for the full power of the system. This results in increased cost and
114
Energies 2021, 14, 3365
negative impact on the efficiency. One of the recent trends in the power electronic studies
is the use of advanced topologies of the DC-DC stage like differential, partial and fractional
power converters that allow operation with lower voltages/currents and minimization of
power losses. The use of these topologies in BES interface is considered below.
Standard Topologies of DC-DC Converters

The choice of the secondary (DC-DC) converter or the converter at the battery end
depends on several factors. First of the all, this converter has to be bidirectional. Besides,
this reduction of the losses requires that the number of switches is minimal, which enables
only simple choppers. Finally, the configuration/mode of the two-stage interface converter
(stabilizing or unfolding) as well as the voltage of the battery are important. Below, the
latter issue is addressed in detail.
In the case of the two-stage converter with a stabilized DC-link, its grid unit may
operate correctly if the voltage of the DC-link is higher than the amplitude of the grid
voltage. On the other hand, keeping this voltage level on the battery is not reasonable
because it would reduce the advantages of the two-stage configuration. Taking into account
the realistic voltage gains of the circuits with the boost function of 2–3 and voltage difference
of the fully discharged and fully charged Li-Ion batteries of 60–100%, the voltage of
the fully charged battery could be at least twice lower than the amplitude of the grid
voltage or about 200 V. In the given paper, this level is considered as a medium level,
but such batteries abbreviated as MV BES. The case of the unfolding configuration/mode
additionally requires that the DC-DC chopper is capable of converting the BES voltage
down to zero level. Together, as a result, the following buck-boost converters are suitable
for both configurations/modes: classical (Figure 18a) and non-inverting (Figure 18b)
bidirectional buck-boost converter, Zeta-SEPIC circuitry that is linked through its primary
inductor to the battery or the grid (Figure 18c), as well as the bidirectional Čuk converter
(Figure 18d). In addition, a synchronous buck converter (Figure 18) is applicable in the case
of the stabilizing configuration/mode only.
Rect./inv./unf.
rect./inv./unf.
To rectifier-
To rectifier-
inverter or
MV BES or
or MV BES
inverter or
unfolder
MV BES
unfolder
MV BES
(a) (b) (c)

To rectifier-
To rectifier-
inverter or
unfolder
MV BES
MV BES
inverter
(d) (e)
Figure 18. DC-DC stage of BES interface converters with two stages: (a) inverting bidirectional buck-boost, (b) bidirectional
non-inverting buck-boost, (c) bidirectional zeta-SEPIC, (d) bidirectional Čuk, (e) synchronous buck.
Zeta-SEPIC-Čuk topologies, however, are usually not considered as powerful con-

verters, which disables them for applications like BESS. On the other hand, the remain-
ing (non)inverting buck-boost and synchronous buck topologies can be equipped with
additional elements for reducing losses, smoothing current ripple and better control per-
115
Energies 2021, 14, 3365
formance (more accurate regulation for the same range of duty cycle): add-on circuits for
zero-current/resonant switching, tapped (coupled) inductors or qZ links [120].
Differential Power Converters

Differential power converters (DPCs) are a kind of partial power processor (PPP). In
turn, PPPs are a quite recent group of converters that are typically used in conjunction
with renewable energy sources and storages. As it follows from their title, the main feature
refers to dealing only with a part of the total system power. PPPs can be systematized
in a number of ways, such as considering their topology or application. However, most
commonly, PPPs can be divided into different groups [124] according to their power flow.
From this point of view, three groups could be differentiated: differential power converters
that internally link elements of the systems, partial power converters connecting system
input and output, fractional power converters dealing with a fraction of entire set of power
sources/storages, as well as mixed topologies. Finally, it must be noted that the difference
between differential, partial and fractional power converters is sometimes quite fuzzy.
For example, in [125,126], identical topologies are entitled as partial and differential. In a
similar way, most of the topologies considered in [127] as partial power converters, in fact,
operate as fractional power converters.
DPCs are mainly used in various balancing systems [128,129]. There are two types of
such converters. The first one transfers the energy between two typically adjacent elements
and is known as element-to-element (E2E) converter; alternatively, in another option, the
energy circulates through a common bus (B2E). The DPC normally operates with batteries,
but according to some reports, this converter technology is applied to photovoltaics [128].
The E2E architecture is used in systems with the same type of cells, for example,
for balancing batteries (Figure 19a). The main advantage of this architecture is that each
converter operates at significantly lower voltage and current values than the entire system.
The disadvantages of this architecture are the interconnectedness of the converters and the
impossibility of their operation separately.
DC/DC Pdk Pdk DC/DC

DC/DC
Common bus
Common bus
Common bus
Virtual bus
Pd1 DC/DC Pd3 Pd3 DC/DC
DC/DC

DC/DC
(a) (b) (c)
Figure 19. Differential power converters: (a) Element-to-Element, (b) Element-to-common-bus, (c) Element-to-virtual-bus.
The B2E architecture works with a common bus connected to the output (Figure 19b)
or with the independent “virtual” bus dedicated to the energy transfer (Figure 19c). Each
element is connected to the bus via own converter. Compared to the E2E architecture, this
approach is more flexible, but neighboring cells are independent of each other. However,
an isolating converter suitable for the full bus voltage is required.
All kinds of DPCs fit well the cell balancing function needed also in BESSs. At the
same time, the use of B2E DPCs as BES and grid interface converters is complicated due
to the following: (1) DC output requires an inverter or unfolder and (2) because the total
power of converters, in fact, is not reduced, but just split into several parts. Lastly, the E2E
DPCs are not applicable as BES and grid interface converters due to the absence of the
common link.
116
Energies 2021, 14, 3365
Partial Power Converters

Another group of PPP links the input and the output of the system. While one
part of the energy from the source to the load goes directly, the converter transfers only
the necessary reminder. As can be seen from Figure 20, in the classical converter type
(Figure 20a), all the output power passes through the converter, which leads to a higher
efficiency and significant losses. In the case of a PPC, a significant part of the total energy
enters the load without conversion and does not produce losses. Only the energy going
through the converter, adjusted by the converter to control the energy flow, produces losses
(Figure 20b). Thus, compared to a classic converter, PPCs have potentially better efficiency
and smaller dimensions for the same power. The PPCs may operate with reduced voltage
(Figure 20c), current or both.
Pin Pout Pin Pconst Pout

Full Power DC/DC Vconv
Converter
Partial Power Vbus
Converter Vbatt
Pppc Pvar
Ploss
Ploss –
(a) (b) (c)
Figure 20. Full power operation vs. partial power operation: (a) power distribution in full power converter, (b) power
distribution in partial power converter, (c) diagram showing operation with reduced voltage.
It is possible to distinguish two groups of PPCs: with an isolated and with a non-
isolated converter. The isolated converter can be applied in a quite free form. That is
why such PPCs can be of two types: parallel input—serial output, as well as serial input—
parallel output (Figure 21). In the first case (Figure 21a), the input source and the input
of the converter are connected in parallel, while the output of the converter and the input
source are connected in series (S-PPC). The configuration is suitable to increase the voltage.
In respect to the battery, the parallel input converters can be considered as partial current
converters because only part of their battery current is transferred to the output (bus)
through the converter. On the other hand, in respect to the output, operation occurs with
reduced voltage because only part of output voltage is applied to the converter (see also
Figure 20c).
+ Vbatt + Isolated DC/DC + + + + Isolated DC/DC + I*conv +

Iconv
Ibus
Vconv V*conv Vconv V*conv
– from battery – Vbus Vbatt – from battery – Vbus

from battery from battery
Ibatt
Vbus=Vbatt +V*conv Vbus =V*conv

Vconv=Vbatt Vbatt Vbus Vbus =Vbatt–Vconv
Ibatt =Ibus+Iconv Ibus=Ibatt +I*conv
– – – –
(a) (b)
Figure 21. Structures of partial power converters: (a) parallel input—series output, (b) parallel output—series input.
In the second case (Figure 21b), the input source is connected in series with the input
of the converter, but its output and the source are connected in parallel (P-PPC). The
configuration is suitable to obtain higher output current. In respect to the battery, the series
117
Energies 2021, 14, 3365
input converters can be considered as partial voltage converters because only part of their
battery voltage is converted and passed to the output.
It is obvious that both topologies are symmetrical and counter-reversible. In respect to
the output (bus), the first configuration is a partial voltage converter, but the second one—a
partial current converter. To some extent, these PPCs are similar to an autotransformer and
can be described by similar mathematical expressions extracted from Kirchhoff’s voltage
and current laws.
The PPC topology provides benefits when the difference between the input and the
output voltage is relatively small and only a small amount of energy is being converted
by PPC. Due to a more complex design and a larger number of active elements, the larger
difference between the input and the output voltages produces lower efficiency. Moreover,
at 100% of the difference, the efficiency will be less than that of a classical converter.
Practical PPC implementation depends on the particular application. Normally, re-
ports consider PPC with a DAB converter at each end of an isolating transformer that
produces a fully bidirectional PPC (Figure 22a). In many applications, the bidirection-
ality can be omitted, but PPC—reasonably simplified. For example, in [130,131], which
address PV systems, the simplification finally produces full-bridge + buck configuration,
in [131,132]—full-bridge + push-pull, but in [130,132]—a kind of classical flyback. The
latest converter can be easily turned to a bidirectional one (Figure 22b).
+ Vbatt Vbus +
+ Vbatt Vbus +
Vconv
Vconv V*conv V*conv
MV BES
MV BES
– –
– –
(a) (b)
Figure 22. Examples of PPCs: (a) parallel input—series output PPC with DAB, (b) parallel input—series output PPC with
flyback converter.
A PPC topology with a non-isolating converter could be potentially simpler, contain
fewer components, and have higher efficiency. Attempts have been made to implement
such non-isolating schemes. For example, refs. [133–135] report the voltage buck-boost
topology based implementation applied for battery or bus voltage magnification. It is
pointed out that the extra feedback capacitor installed in these schematics is required for
direct power feedforwarding. However, it is possible to show that the obtained converters
are, in fact, ordinary boost or buck converters—see [136] for details.
Fractional Power Converters

Fractional Power Converters (FPCs) deal with an explicit part of the entire power
supply, for example, with several cells of BES [137,138]. In contrast to PPCs, where the
reduced operating voltage of the converter is obtained as a difference on the entire in-
put/output, FPCs process already reduced voltage—a section of the entire power supply
(similar consideration could be applied to current conversion).
The fractional power processing may utilize an isolated (Figure 23b) or a non-isolated
(Figure 23c) scheme. Successful examples of non-isolated converter use have been demon-
strated in conjunction with a battery of PEVs [137] or grid BESS [138]. The mentioned
reports explore several DC-DC choppers functioning as FPC (Figure 24 shows discharging
configurations in black, but charging—in gray). In this case, the non-isolated converter
obviously deals with reduced voltage, thus providing true partial power processing. On
118
Energies 2021, 14, 3365
the other hand, the fraction of the power supply associated with the converter operates dif-
ferently from the rest of BES. It has different average charge/discharge current. Moreover,
depending on the applied chopper, it may conduct pulse-mode current. This may lead to
shorter operation cycles, limited state of charge and, finally, may lead to a worse state of
health for the “processed” cells.
Vbatt(f2)=Vconv
Vbus =Vbatt(f1)+V*conv
Vbus =Vbatt(f1)+V*conv
+ + + + + + + + + +
Vconv DC/DC V*conv DC/DC V*conv DC/DC –V*conv
– – – – –
Vbatt Vbus Vbatt(f2)=Vconv
Vbatt(f1)
Vbatt(f1)
– – – –
(a) (b) (c)
Figure 23. Partial power converters (a) vs. fractional power converters: (b) isolated, (c) non-isolated.
+ + + + + +
+ Vbatt(f2)=Vconv V*conv Vbatt(f2)=Vconv – Vbatt(f2)=Vconv
Vbus=Vbatt(f1)+V*conv
+ V *conv
V*conv
– – +
– –
–
+ + +
Vbatt(f1) Vbatt(f1) Vbatt(f1)
– – –
– – –
(a) (b) (c)
Figure 24. Examples of non-isolated fractional power converters (in discharge mode): (a) buck, (b) boost, (c) buck-boost.
The FPC shown in Figure 24c is also quite impractical because the polarities of the
input/output voltage are different, which splits the battery or narrows the regulation range.
On the other hand, the use of non-inverting buck-boost topology would double the static
and dynamic losses of the switches.
Finally, the considerations on the non-isolated PPC with a feedback capacitor (given
in the previous section) may also produce, in fact, an FPC if the feedback capacitor is sub-
stituted with an energy source or storage (battery, supercapacitor, PV cells etc.) capable of
keeping its voltage at a constant level. Then the part of the current is actually bypassed, but
the other—processed in the converter (Figure 25a). Practical importance of this converter
is questionable because one fraction of the battery is loaded with increased current and
charged/discharged more intensively.
MV BES (fraction 2)
– + Ibypass
Vbatt(f2)
+ + + +
+
(fraction 1)
Ifraction Vconv V*conv Vbus

Vbus=V*conv
MV BES
– –
Vbatt(f1)=Vconv +
Vbatt
– –
– –
(a) (b)
Figure 25. Specific kinds of FPC: (a) FPC synthesized from “non-isolated” synchronous buck PPC, (b) fractional power
conversion with “virtual fraction” [139].
119
Energies 2021, 14, 3365
Another example of the implementation of non-isolated partial power conversion

is given in [139]. This example, however, may also be considered as a fractional power
converter with “virtual” fraction formed of a DAB and an ordinary capacitor. Here,
the DAB processes low voltage of the capacitor while the battery is attached in series
without processing (Figure 25b). Although, in [139], with the focus on the DC-AC systems,
partial conversion of power occurs in the DC-link. Due to the limited energy capacity,
the configuration is suitable for compensation of regular short term voltage fluctuation
or current compensation that happens, for example, within the cycle of the supply grid
(20 ms).
4.4.2. Two-Stage Converters with Pulsating DC-Link

In order to reduce the overall switching losses of the two-stage system, a configuration
with inverting unfolder can be used. In this case, the DC-DC converter forms unipolar sine
half-waves in the pulsating DC-link, but the interface inverter applies it to the grid with
the proper polarity.
The Li-ion battery can handle the current ripple without significant effect on their
lifetime, thus the use of pulsating current can be justified [140,141].
1-ph Unfolders
As it was stated previously, an unfolding circuit provides grid-frequency commutation
of the unipolar voltage formed by a high-frequency switch mode DC-DC converter to
provide sinewave matching to the grid polarity. Paper [142] proposes a combination
of a buck/boost non-inverting converter and an unfolding H-bridge (Figure 26a). This
configuration directly corresponds to Figure 17b and can be considered as a standard double
stage converter with a pulsating DC-link. In [143], the operation and the experimental
verification of a buck-boost inverter/converter based on tapped inductor are addressed.
The inductor magnetically couples four windings with equal turn-ratio (Figure 26b). In the
converter presented in [143], in contrast to [142], explicit parts cannot be identified.
HV or MV Half-wave shaper Polarity converter

BES Grid
VGR(t)
t
VDC(t)
VBT(t)
t
t
HV or MV
Grid
Gri
G iid BES
(a) (b)
Figure 26. Examples of single-phase unfolders: (a) buck-boost converter with unfolding H-bridge, (b) tapped inductor
buck-boost converter.
3-ph Unfolders
Three-phase converters with low frequency unfolding stage utilize principles similar
to those of single-phase unfolders. However, the presence of three phases requires that at
least two voltages/currents be formed actively by the dedicated voltage/current source
(shaper), but the third one is obtained as a sum/difference of the other two. Within a period
of the grid, the principles how the actively shaped voltages/currents are applied to the
grid change six times (Figure 27a,b).
Working principles of unfolder topologies are provided in [144]; however, the con-
verters described there are unidirectional and do not fit the requirements of bidirectional
operation. An example of such converter from [144] is a topology derived from a three-
phase two-level voltage source inverter. In this case, amplitude modulated high-frequency
120
Energies 2021, 14, 3365
output of a phase modulated high-frequency inverter (H-bridge was taken as an example)

is rectified and filtered. Then the output of a filter is unfolded by a three-phase inverter,
thus forming three-phase alternating current. For the bidirectional operation, one or mul-
tiple DC-DC converters should be used as current sources. For example, refs. [145,146]
show a three-phase inverter, where two DC-DC converters were taken as current sources
and are connected in parallel to the BES. DC-DC converter outputs are connected in series,
thus forming three voltage levels—high, low and neutral. Then the modulated voltage
waveforms are unfolded by a three-level inverter, which is derived from a diode clamped
multilevel converter [146] or Vienna rectifier [145] (Figure 27c). The study in [147] provides
experimental verification of the topology in [146].
i/v(t) va(t) va(t) vb(t) vc(t)

i+,v+(t) HV or MV
Commutation matrix
BES Grid
i,vup (t)
t DC/DCup va(t)
t vb(t) vb(t)
ifl,vfl(t)
i,vbt(t)
i,vbt(t) i,vup (t) vc(t)

t vc(t) DC/DCbt
i-,v-(t) t
(a) (b) (c)
Figure 27. Three-phase unfolders: (a) general principle, (b) operation diagram, (c) example of implementation.
5. Generalizations and Discussion

One of the main trends in the area of residential BESSs is a shift from stand-alone
systems for a particular task, like the increasing local renewable energy self-consumption
using basic algorithms, to a complex integrated system, aiming to achieve multiple different
goals. Some commercial BESSs can provide a supply during a power cut, effectively acting
as uninterruptable power supplies. The use of advanced energy management algorithms
enables smart scheduling and energy trading, further improving the functionality of these
systems. The integration on a system level allows provision of ancillary grid services and
enables forming of a microgrid with extended levels of flexibility. Nevertheless, currently
there are still a number of technological and legal barriers that limit a ubiquitous application
of the residential BESS. However, the new regulations, subsidies and initiatives offered
and developed by governmental structures in many countries are aimed to stimulate the
installation of these systems and make full use of their potential [148].
Our review of the commercially available BESS and the corresponding intellectual
property right items shows that despite a whole range of available solutions, the market of
residential BESSs is still advancing. Presently, it is strongly influenced by a forecasted price
reduction of the Li-ion battery cells and further improvements in the battery chemistry.
Technical information about available products is rather limited. Most of the BESSs utilize
a low-voltage (around 50 V) Li-ion battery, which results in high current and requires
the use of a transformer. Most likely, two-stage converters with galvanic isolation are
used, limiting the overall efficiency of the system. The typical BESSs have the efficiency
below 90%, whereas power electronic converters could be responsible for around half of
power losses in the system. On the other hand, the low voltage batteries require simpler
battery management and protection systems, making them less expensive and more feasible
commercially. Because of these compromises, the current BESSs have strict limitations
on the efficiency due to the use of low voltage batteries that are associated with high
currents and more complex converter topologies. On the other hand, the batteries with
increased voltages would enable the use of non-isolated topologies with potentially higher
efficiencies. Several companies have already introduced such products to the market.
121
Energies 2021, 14, 3365
The typical power electronic interface of a battery with the grid is based on a two-stage
configuration, comprised of a bidirectional DC-DC converter and a DC-AC inverter/rectifier
connected via an intermediate DC-link. Modern Li-ion batteries can sustain current ripples
associated with the grid frequency very well, even in single-phase systems. It is therefore
possible to connect a battery of sufficiently high voltage to an inverter directly. Still, due
to variation of battery voltage depending on its state of charge, the efficiency and power
quality of such system is compromised. As a result, the intermediate DC-DC stage is still
necessary to stabilize the DC voltage and obtain better performance.
With the possibility of using non-isolated converters to interface HV batteries, the
standard approach would assume application of well-known DC-DC topologies, like
buck, boost and buck-boost together with the grid inverter stage. This makes both of the
two conversion stages process full power and exhibit high switching frequency, which
still compromises the efficiency. One of the approaches that is widely addressed in recent
studies is to use emerging solutions, like PPCs and FPCs at the DC-DC stage, which, as it
is already reported, have been achieved extremely high efficiency values. However, the
practical aspects, including transient operating modes, protection and cost, need to be
evaluated further to justify this technology.
A range of alternative concepts utilizes a pulsating DC-link instead of the stabilized
one. This brings the converter system closer to the single-stage converter, where only
DC-DC stage operates with high frequency, while the grid-side inverter just unfolds
the unipolar pulses into the sine wave and exhibits conduction losses only. A similar
approach can be applicable to both single- and three-phase systems [145]. In addition,
mixed concepts with fluctuating DC-link were also proposed, aiming to distribute losses
more evenly between the stages [149]. On the other hand, it would be much more difficult
to integrate other sources into such DC-link and therefore such solutions are generally
suitable for AC-coupled BESS only.
Impedance source inverters are another group of topologies that allow voltage pre-
regulation at a “virtual DC-link” before it is inverted into a sinusoidal waveform. Single-
phase, three-phase and three-level configurations of these inverters were proposed in [150].
They can be more short-circuit-proof, as the shoot-through state is one of the inherent
operating modes of such topologies. However, some studies show that the voltage
stress on semiconductors and volume of components can be larger than for the stan-
dard two-stage configurations [151]. Moreover, only few studies address bidirectional
operation of impedance-source converters [152,153].
In conclusion, there is a range of solutions for HV BESSs that are potential alternatives
to standard buck-boost plus inverter configuration. The most optimal choice would
evidently depend on the parameters of the system and its configuration. For the systems
that incorporate a DC bus for integration of renewables and loads, a PPC/FPC with a
bridge-type bidirectional inverter/rectifier seems to be a very promising solution. On the
other hand, for an AC-coupled BES, the use of pulsating/fluctuating DC-link and unfolding
inverter can bring an advantage in terms of switching loss and absence of a bulky capacitor.
Still, the behavior of such configurations in practical applications, including transient
modes and fault ride-through capabilities, needs to be addressed in more detail.
The configurations that include multilevel inverter topologies also seem quite promis-
ing for residential BESs. Despite generally being used in high-power applications, there
are successful commercialization examples of this technology in residential applications.
Recent works aim to bring such inverters on a new level, particularly taking advantage
of developments in WBG semiconductors [115,116]. The systems with multilevel topolo-
gies potentially enable the use of battery stacks with lower voltage levels as compared to
standard two-level inverters. This could result in a more optimal storage configuration.
On the other hand, presently, the cost of WBG devices is still relatively high to make such
multilevel inverters feasible in commercial BESSs.
The comparative analysis of evaluated power electronic interfaces is presented in
Tables 3 and 4. The considerations above show that the most promising units of composite
122
Energies 2021, 14, 3365
BESS grid interface converters have somehow completing features (see Table 3 for details).
For example, unfolding circuits provide neither DC regulation at the corresponding port
nor AC half wave forming. This functionality, however, can be performed by a DC-DC
converter. Multilevel converters without pulse mode control do not provide pulse mode
regulation between levels, but partial power converters—provide regulation within a
narrow range. Besides, the multilevel converters and unfolding units have no switching
losses, but have significant conduction losses (Table 3). At the same time, the partial power
converters can reduce both. A logical conclusion from the above is to combine the units
with the adjacent features (Table 4).
Table 3. Losses of converters and energy conversion principles in BESS grid interface.
Stage Main Function Peculiarities

Full Power Switch-Mode + Established technology,
Rectifiers/Inverter Forming AC
(origin for comparison) − High voltage input, high switching frequency, bulky filter
Full Power Switch-Mode + Established technology, wide regulation range,
Forming DC
DC/DC Converters − Full power operation, high switching frequency
+ Operation with part of rated power
Partial Power Converters Forming DC
− Developing technology, limited regulation range
+ Established technology, small grid filter
Multilevel Converters Forming AC or DC
− Control and hardware complexity
+ No switching losses
Unfolding Circuits Commutation
− Developing technology, no regulation
Table 4. Promising combination of converters to form BESS grid interface.
Configuration Advantages Disadvantages

Lower efficiency at most of the
Single stage DC-AC Bidirectional Max. efficiency at a particular operation points,
Inverters/Rectifiers operation point Minimal battery voltage > amplitude
of grid voltage
Voltage stress on semiconductors and
Impedance-Source Bidirectional Battery voltage pre-regulation volume of components is larger
Inverters/Rectifiers Short-Circuit Proof Complicated bidirectional operation
Developing technology
Higher efficiency at the most of
operation points, Lower maximal efficiency,
Bidirectional inverter/ Wide battery voltage range,
rectifier + Full Power DC-DC Both stages operate at full power and
Allows integration of renewables high switching frequency
into DC-link
operation points,
Bidirectional inverter/ Narrow battery voltage range,
Allows integration of renewables into
rectifier + PPC DC-DC Developing technology
DC-link, DC-DC operates with part of
rated power
Low grid filter size and volume,
High component count
Multilevel DC-DC and DC-AC Utilization of low voltage semiconductors,
Complex control
Modular design
operation points,
No integration of renewables
Unfolder + Full Power DC-DC Wide battery voltage range,
No switching losses in grid stage, into DC-link
No DC-link capacitor
6. Conclusions and Future Trends

This paper gives an insight into the field of storage systems for residential applications
together with associated technologies and developments. To provide a broader view, the
123
Energies 2021, 14, 3365
current state of these systems is addressed from multiple directions, including battery
technologies, their market, standards, and grid interface converters.
Instigated by the on-going paradigm shift from centralized to distributed power
generation, the storage technologies will become one of the key components of the future
electrical grids that enable more optimal use of the conventional and local renewable
energy sources and ensure the power supply security. However, a range of technological
and regulatory barriers still stand in the way of these systems, limiting their benefits
and potential.
Today’s market for dedicated residential storage systems is still in the process of
being established. It is currently very dynamic, and several manufacturers have already
introduced and commercialized their solutions, with more companies and products being
announced and trying to enter the market every year. Still, the price for residential solutions
is relatively high for a private client, while the return of investment is not evident in
many cases.
The developments and price reduction of Li-ion battery technologies are mainly driven
by massive transportation electrification and this trend will continue in the following years.
Despite the distinct potential of vehicle to grid (V2G) solutions, they are unlikely to be able
to replace stationary battery systems and their functions due to economic reasons, mainly
related to lifetime and cycle-cost. Nevertheless, the use of second-life Li-ion batteries for
stationary storage has certain potential.
Batteries based on the Li-ion technology are currently dominating the market, however,
at a certain point, the price and performance of other battery technologies, like flow batter-
ies, is likely to make them a more expedient choice for larger-scale stationary solutions.
According to our analysis, the majority of commercial residential storages are currently
using low voltage batteries with voltages of around 50 V, mainly due to the cheaper price
per kWh. These batteries are typically interfaced with the grid by means of a power
electronic converter with a transformer to provide required voltage matching and galvanic
isolation. However, the mass production of HV EV batteries along with their second-life
use is likely to make the HV stationary storage solutions more popular in the residential
sector. This would make the use of non-isolated interface converter topologies attractive
due to their typically lower component count and higher efficiency. In addition to standard
and typically used topologies, like buck-boost or bridge, which are rated for full power of
the system, the recent research interest is also focused on partial- (fractional-, differential-)
power converters. Such topologies have the potential to offer even further improvement of
efficiency in various operating conditions.
Presently, many countries are introducing initiatives that are either directly (by subsi-
dies) or indirectly (via marginal feed-in tariffs) encouraging the use of local energy storage.
Moreover, a range of standards is being developed to regulate the use of such systems
and facilitate unleashing of their full potential. In addition to basic renewable energy
self-consumption increase, battery-based storage systems can provide uninterruptable
power supply functionality, offer ancillary grid service support, enable peer-to-peer energy
trading etc. Together with the large-scale global investments in the battery technologies it
is highly likely that in the following decades, the residential battery systems will follow the
route of photovoltaics and become an essential and inherent part of the future power grid.
Author Contributions: Conceptualization, I.A.G. and A.B. (Andrei Blinov); investigation of BESS
market solutions, R.S.; investigation of IP right items and BESS market state for distribution grids,
M.V.; topological analysis of single stage pulse mode converters, I.A.G.; topological analysis of
multilevel converters, A.B. (Alexander Bubovich); topological analysis of partial power processors,
I.A.G. and R.S.; supervision, I.A.G.; Validation, I.A.G.; writing—original draft, all; Supervision and
editing—A.B. (Andrei Blinov) and D.P. All authors have read and agreed to the published version of
the manuscript.
Funding: This work was supported in part by the European Economic Area (EEA) and Norway
Financial Mechanism 2014–2021 under Grant EMP474 and in part by the Estonian Research Council
grant (PRG1086).
124
Energies 2021, 14, 3365

Data Availability Statement: Data is contained within the article.
References
1. The European Parliament and Council of the European Union. Directive 2009/28/EC on the promotion of the use of energy from
renewable sources. Off. J. Eur. Union 2009, L140, 16–62.
2. The European Parliament and Council of the European Union. Directive (EU) 2018/2001 on the promotion of the use of energy
from renewable sources. Off. J. Eur. Union 2018, L328, 82–209.
3. European Commission. Energy Strategy. Available online: https://ec.europa.eu/energy/topics/energy-strategy-and-energy-
union_en (accessed on 13 November 2020).
4. Andrijanovits, A.; Hoimoja, H.; Vinnikov, D. Comparative Review of Long-Term Energy Storage Technologies for Renewable
Energy Systems. Elektron. Elektrotechnika 2012, 118, 21–26. [CrossRef]
5. Vinnikov, D.; Hoimoja, H.; Andrijanovits, A.; Roasto, I.; Lehtla, T.; Klytta, M. An improved interface converter for a medium-
power wind-hydrogen system. In Proceedings of the 2009 International Conference on Clean Electrical Power, Capri, Italy, 9–11
June 2009; Institute of Electrical and Electronics Engineers (IEEE): New York, NY, USA, 2009; pp. 426–432. [CrossRef]
6. Liang, M.; Liu, Y.; Xiao, B.; Yang, S.; Wang, Z.; Han, H. An analytical model for the transverse permeability of gas diffusion
layer with electrical double layer effects in proton exchange membrane fuel cells. Int. J. Hydrog. Energy 2018, 43, 17880–17888.
[CrossRef]
7. Liang, M.; Fu, C.; Xiao, B.; Luo, L.; Wang, Z. A fractal study for the effective electrolyte diffusion through charged porous media.
Int. J. Heat Mass Transf. 2019, 137, 365–371. [CrossRef]
8. Vinnikov, D.; Andrijanovits, A.; Roasto, I.; Jalakas, T. Experimental study of new integrated DC/DC converter for hydrogen-based
energy storage. In Proceedings of the 2011 10th International Conference on Environment and Electrical Engineering, Rome, Italy,
8–11 May 2011; Institute of Electrical and Electronics Engineers (IEEE): New York, NY, USA, 2011; pp. 1–4.
9. Stecca, M.; Elizondo, L.R.; Soeiro, T.; Bauer, P.; Palensky, P. A Comprehensive Review of the Integration of Battery Energy Storage
Systems into Distribution Networks. IEEE Open J. Ind. Electron. Soc. 2020, 1, 46–65. [CrossRef]
10. Wang, G.; Konstantinou, G.; Townsend, C.D.; Pou, J.; Vazquez, S.; Demetriades, G.D.; Agelidis, V.G. A Review of Power
Electronics for Grid Connection of Utility-Scale Battery Energy Storage Systems. IEEE Trans. Sustain. Energy 2016, 7, 1778–1790.
[CrossRef]
11. Pires, V.F.; Romero-Cadaval, E.; Vinnikov, D.; Roasto, I.; Martins, J. Power converter interfaces for electrochemical energy storage
systems–A review. Energy Convers. Manag. 2014, 86, 453–475. [CrossRef]
12. Yao, Z. Review of Dual-Buck Type Single-Phase Grid-Connected Inverters. IEEE J. Emerg. Sel. Top. Power Electron. 2020, 1.
[CrossRef]
13. Available online: https://nakedsolar.co.uk/storage/ (accessed on 7 June 2021).
14. Available online: https://www.which.co.uk/reviews/solar-panels/article/solar-panels/solar-panel-battery-storage-a2AfJ0s5
tCyT (accessed on 7 June 2021).
15. Saez-De-Ibarra, A.; Laserna, E.M.; Stroe, D.-I.; Swierczynski, M.J.; Rodriguez, P. Sizing Study of Second Life Li-ion Batteries for
Enhancing Renewable Energy Grid Integration. IEEE Trans. Ind. Appl. 2016, 52, 4999–5008. [CrossRef]
16. Rezania, R.; Prüggler, W. Business models for the integration of electric vehicles into the Austrian energy system. In 2012 9th
International Conference on the European Energy Market; Institute of Electrical and Electronics Engineers (IEEE): New York, NY, USA,
2012; pp. 1–8.
17. Robert, S. Introduction to Batteries—An IEEE Course; IEEE: New York, NY, USA, 2013.
18. Hu, X.; Zou, C.; Zhang, C.; Li, Y. Technological Developments in Batteries: A Survey of Principal Roles, Types, and Management
Needs. IEEE Power Energy Mag. 2017, 15, 20–31. [CrossRef]
19. Electricity Storage and Renewables: Costs and Markets to 2030; The International Renewable Energy Agency (IRENA): Abu Dhabi,
United Arab Emirates, 2017; pp. 36–49.
20. Cao, J.; Emadi, A. Batteries Need Electronics. IEEE Ind. Electron. Mag. 2011, 5, 27–35. [CrossRef]
21. Ferreira, B. Batteries, the New Kids on the Block. IEEE Power Electron. Mag. 2019, 6, 32–34. [CrossRef]
22. Chen, C.; Plunkett, S.; Salameh, M.; Stoyanov, S.; Al-Hallaj, S.; Krishnamurthy, M. Enhancing the Fast Charging Capability of
High-Energy-Density Lithium-Ion Batteries: A Pack Design Perspective. IEEE Electrif. Mag. 2020, 8, 62–69. [CrossRef]
23. Fulli, G.; Masera, M.; Spisto, A.; Vitiello, S. A Change is Coming: How Regulation and Innovation Are Reshaping the European
Union’s Electricity Markets. IEEE Power Energy Mag. 2019, 17, 53–66. [CrossRef]
24. Lukic, S.; Emadi, A. Charging ahead. IEEE Ind. Electron. Mag. 2008, 2, 22–31. [CrossRef]
25. Khaligh, A.; Li, Z. Battery, Ultracapacitor, Fuel Cell, and Hybrid Energy Storage Systems for Electric, Hybrid Electric, Fuel Cell,
and Plug-In Hybrid Electric Vehicles: State of the Art. IEEE Trans. Veh. Technol. 2010, 59, 2806–2814. [CrossRef]
26. Quiros-Tortos, J.; Ochoa, L.; Butler, T. How Electric Vehicles and the Grid Work Together: Lessons Learned from One of the
Largest Electric Vehicle Trials in the World. IEEE Power Energy Mag. 2018, 16, 64–76. [CrossRef]
125
Energies 2021, 14, 3365
27. Chen, N.; Ma, J.; Li, M.; Wang, M.; Shen, X.S. Energy Management Framework for Mobile Vehicular Electric Storage. IEEE Netw.
2019, 33, 148–155. [CrossRef]
28. Chandler, S.; Gartner, J.; Jones, D. Integrating Electric Vehicles with Energy Storage and Grids: New Technology and Specific
Capabilities Spur Numerous Applications. IEEE Electrif. Mag. 2018, 6, 38–43. [CrossRef]
29. Al-Rubaye, S.; Al-Dulaimi, A.; Ni, Q. Power Interchange Analysis for Reliable Vehicle-to-Grid Connectivity. IEEE Commun. Mag.
2019, 57, 105–111. [CrossRef]
30. Arboleya, P.; Bidaguren, P.; Armendariz, U. Energy Is on Board: Energy Storage and Other Alternatives in Modern Light Railways.
IEEE Electrif. Mag. 2016, 4, 30–41. [CrossRef]
31. Sinhuber, P.; Rohlfs, W.; Sauer, D.U. Study on power and energy demand for sizing the energy storage systems for electrified
local public transport buses. In 2012 IEEE Vehicle Power and Propulsion Conference; Institute of Electrical and Electronics Engineers
(IEEE): New York, NY, USA, 2012; pp. 315–320. [CrossRef]
32. Guarnieri, M.; Morandin, M.; Ferrari, A.; Campostrini, P.; Bolognani, S. Electrifying Water Buses: A Case Study on Diesel-to-
Electric Conversion in Venice. IEEE Ind. Appl. Mag. 2017, 24, 71–83. [CrossRef]
33. Sorensen, A.J.; Skjetne, R.; Bo, T.; Miyazaki, M.R.; Johansen, T.A.; Utne, I.B.; Pedersen, E. Toward Safer, Smarter, and Greener
Ships: Using Hybrid Marine Power Plants. IEEE Electrif. Mag. 2017, 5, 68–73. [CrossRef]
34. Paul, D. A History of Electric Ship Propulsion Systems [History]. IEEE Ind. Appl. Mag. 2020, 26, 9–19. [CrossRef]
35. Vicenzutti, A.; Bosich, D.; Giadrossi, G.; Sulligoi, G. The Role of Voltage Controls in Modern All-Electric Ships: Toward the all
electric ship. IEEE Electrif. Mag. 2015, 3, 49–65. [CrossRef]
36. Roboam, X.; Sareni, B.; De Andrade, A. More Electricity in the Air: Toward Optimized Electrical Networks Embedded in
More-Electrical Aircraft. IEEE Ind. Electron. Mag. 2012, 6, 6–17. [CrossRef]
37. Misra, A. Energy Storage for Electrified Aircraft: The Need for Better Batteries, Fuel Cells, and Supercapacitors. IEEE Electrif.
Mag. 2018, 6, 54–61. [CrossRef]
38. Crittenden, M. Ultralight batteries for electric airplanes. IEEE Spectr. 2020, 57, 44–49. [CrossRef]
39. Manz, D.; Piwko, R.; Miller, N. Look before You Leap: The Role of Energy Storage in the Grid. IEEE Power Energy Mag. 2012, 10,
75–84. [CrossRef]
40. Farrokhabadi, M.; Solanki, B.V.; Canizares, C.A.; Bhattacharya, K.; Koenig, S.; Sauter, P.S.; Leibfried, T.; Hohmann, S. Energy
Storage in Microgrids: Compensating for Generation and Demand Fluctuations While Providing Ancillary Services. IEEE Power
Energy Mag. 2017, 15, 81–91. [CrossRef]
41. Cagnano, A.; De Tuglie, E.; Mancarella, P. Microgrids: Overview and guidelines for practicalimplementations and operation.
Appl. Energy 2020, 258, 114039. [CrossRef]
42. Torres-Moreno, J.L.; Gimenez-Fernandez, A.; Perez-Garcia, M.; Rodriguez, F. EnergyManagement Strategy for Micro-Grids with
PV-Battery Systems and Electric Vehicles. Energies 2018, 11, 522. [CrossRef]
43. Lezynski, P.; Szczesniak, P.; Waskowicz, B.; Smolenski, R.; Drozdz, W. Design andimplementation of a fully controllable
cyber-physical system for testing energy storage systems. IEEE Access 2019, 7, 47259–47272. [CrossRef]
44. Restrepo, C.; Salazar, A.; Schweizer, H.; Ginart, A. Residential Battery Storage: Is the Timing Right? IEEE Electrif. Mag. 2015, 3,
14–21. [CrossRef]
45. González, I.; Calderón, A.J.; Portalo, J.M. Innovative Multi-Layered Architecture forHeterogeneous Automation and Moni-toring
Systems: Application Case of a Photovoltaic SmartMicrogrid. Sustainability 2021, 13, 2234. [CrossRef]
46. James, G.; Peng, W.; Deng, K. Managing Household Wind-Energy Generation. IEEE Intell. Syst. 2008, 23, 9–12. [CrossRef]
47. Duryea, S.; Islam, S.; Lawrance, W. A battery management system for stand-alone photovoltaic energy systems. IEEE Ind. Appl.
Mag. 2001, 7, 67–72. [CrossRef]
48. Lu, X.; Wang, J. A Game Changer: Electrifying Remote Communities by Using Isolated Microgrids. IEEE Electrif. Mag. 2017, 5,
56–63. [CrossRef]
49. Zhong, Q.-C.; Wang, Y.; Ren, B. Connecting the Home Grid to the Public Grid: Field Demonstration of Virtual Synchronous
Machines. IEEE Power Electron. Mag. 2019, 6, 41–49. [CrossRef]
50. Ma, Z.; Pesaran, A.; Gevorgian, V.; Gwinner, D.; Kramer, W. Energy Storage, Renewable Power Generation, and the Grid: NREL
Capabilities Help to Develop and Test Energy-Storage Technologies. IEEE Electrif. Mag. 2015, 3, 30–40. [CrossRef]
51. Rodriguez-Diaz, E.; Chen, F.; Vasquez, J.C.; Guerrero, J.M.; Burgos, R.; Boroyevich, D. Voltage-Level Selection of Future Two-Level
LVdc Distribution Grids: A Compromise Between Grid Compatibiliy, Safety, and Efficiency. IEEE Electrif. Mag. 2016, 4, 20–28.
[CrossRef]
52. IEC TC 120 Group BESS Systems Standardization Plan. Available online: https://assets.iec.ch/public/miscfiles/sbp/120.pdf
(accessed on 28 February 2021).
53. UL STD 1741. Inverters Converters and Controllers for Use in Independent Power Systems; IEEE: New York, NY, USA, 2018.
54. IEC TS 62933-5-1. Electrical Energy Storage (EES) Systems-Part 5-1: Safety Considerations for Grid-Integrated EES Systems-General
Specification; IEEE: New York, NY, USA; ISO/IEC: Geneva, Switzerland, 2017.
55. IEC T R 62543. High-Voltage Direct Current (HVDC) Power Transmission Using Voltage Sourced Converters (VSC); ISO/IEC: Geneva,
Switzerland, 2017.
56. Standard IEC TR 61850-90-7. Communication Networks and Systems for Power Utility Automation-Part 90-7: Object Models for Power
Converters in Distributed Energy Resources (DER) Systems; ISO/IEC: Geneva, Switzerland, 2013.
126
Energies 2021, 14, 3365
57. Standard IEC 62920. Photovoltaic Power Generating Systems-EMC Requirements and Test Methods for Power Conversion Equipment;
ISO/IEC: Geneva, Switzerland, 2017.
58. Standard IEC 62909-1. Bi-Directional Grid Connected Power Converters-Part 1: General Requirements; ISO/IEC: Geneva,
Switzerland, 2017.
59. Standard IEC 62040-5-3. Uninterruptible Power Systems (UPS)-Part 5-3: DC Output UPS-Performance and Test Requirements; ISO/IEC:
Geneva, Switzerland, 2016.
60. Standard IEC 62040-4. Uninterruptible Power Systems (UPS)-Part 4: Environmental Aspects-Requirements and Reporting; ISO/IEC:
Geneva, Switzerland, 2013.
61. Standard IEC62040-3. Uninterruptible Power Systems (UPS)-Part 3: Method of Specifying the Performance and Test Requirements;
ISO/IEC: Geneva, Switzerland, 2011.
62. The EMerge Alliance Data/Telecom Center Standard Creates an Integrated, Open Platform for Power, Infrastructure, Periph-
eral Device and Control Applications to Facilitate the Hybrid Use of AC and DC Power within Data Centers and Telecom
Central Offices. Available online: https://www.emergealliance.org/standards/data-telecom/standard-faqs/ (accessed on
28 February 2021).
63. Tesla Powerwall Review. Available online: https://www.cleanenergyreviews.info/blog/tesla-powerwall-2-solar-battery-review
(accessed on 25 April 2021).
64. Sonnen Documentation. Available online: http://www.sonnensupportaustralia.com.au/documentation.html (accessed on
25 April 2021).
65. Available online: https://store.enphase.com/storefront/en-us/pub/media/productattach/e/n/envoys-ds-en-us.pdf (accessed on
7 June 2021).
66. Available online: https://store.enphase.com/storefront/en-us/pub/media/productattach/e/n/enphase_encharge_3_datasheet.
pdf (accessed on 7 June 2021).
67. Available online: https://store.enphase.com/storefront/en-us/pub/media/productattach/e/n/encharge_10_datasheet.pdf
(accessed on 7 June 2021).
68. Victron Energy Blue Power. Available online: https://www.victronenergy.com/upload/documents/Brochure-Energy-Storage-
EN_web.pdf (accessed on 25 February 2021).
69. ADARA Power Commercial Energy Storage System. Available online: http://www.adarapower.com/home/commercial-energy-
storage-system/ (accessed on 25 February 2021).
70. Adara Power Introduces 20-kWh Residential Energy Storage Solution. Available online: https://www.solarpowerworldonline.
com/2017/04/adara-power-introduces-20-kwh-residential-energy-storage-solution/ (accessed on 25 February 2021).
71. Schnoder Electric Hybrid Inverter/Charger XW+. Available online: https://solar.schneider-electric.com/wp-content/uploads/
2020/08/DS20200812_XW-120-240-V.pdf (accessed on 25 February 2021).
72. Sunverge Energy AC-Coupled Solar Integration System (SIS). Available online: https://cdn2.hubspot.net/hubfs/2472485/
WebsiteContent/Sunverge_ACSIS_NA_12092016.pdf?t=1485218396447 (accessed on 25 February 2021).
73. LG Home Battery RESU. Available online: https://www.lgessbattery.com/eu/main/main.lg (accessed on 25 April 2021).
74. Solax Triple Power Battery–LFP. Available online: https://www.solaxpower.com/triple-power-battery/ (accessed on
25 April 2021).
75. StorEdge™ On-Grid Solution. Available online: https://www.solaredge.com/solutions/self-consumption#/ (accessed on
25 April 2021).
76. 4 kWh Powervault Lead-Acid Solar Energy Storage Systems. Available online: https://www.ecopowersupplies.com/4kwh-
powervault-lead-acid-solar-energy-storage-systems (accessed on 25 February 2021).
77. Purestorage, I.I. Available online: https://www.puredrive-energy.co.uk/ (accessed on 25 April 2021).
78. Duracell Energy Bank. Available online: https://www.duracellenergybank.com/ (accessed on 25 April 2021).
79. Enphase Encharge 3. Available online: https://store.enphase.com/storefront/en-us/enphase_encharge_3 (accessed on 25 April 2021).
80. Enphase Encharge 10. Available online: https://store.enphase.com/storefront/en-us/enphase_encharge_10 (accessed on
25 April 2021).
81. Available online: https://www.eaton.com/content/eaton/gb/en-gb/catalog/energy-storage/xstorage-home.html/ (accessed on
25 April 2021).
82. xStorage Home. Available online: https://www.samsungsdi.com/ess/index.html (accessed on 25 April 2021).
83. VARTA Pulse Neo. Available online: https://www.varta-ag.com/en/consumer/product-categories/energy-storage-systems/
varta-pulse-neo (accessed on 25 April 2021).
84. Sunny Boy Storage 2.5. Available online: https://www.sma.de/en/products/battery-inverters/sunny-boy-storage-25.html
(accessed on 25 April 2021).
85. EasySolar. Available online: https://www.victronenergy.ru/inverters-chargers/easysolar (accessed on 25 April 2021).
86. EssProTM-Battery Energy Storage. The Power to Control Energy. Available online: https://new.abb.com/docs/librariesprovider7
8/eventos/jjtts-2017/presentaciones-peru/(dario-cicio)-bess---battery-energy-storage-system.pdf?sfvrsn=2 (accessed on
28 February 2021).
87. Battery Energy Storage Solutions for Stable Power Supply. Available online: https://www.nidec-industrial.com/markets/
renewable-energy/battery-energy-storage-solutions/ (accessed on 28 February 2021).
127
Energies 2021, 14, 3365
88. Kheraluwala, M.; Gascoigne, R.; Divan, D.; Baumann, E. Performance characterization of a high-power dual active bridge
DC-to-DC converter. IEEE Trans. Ind. Appl. 1992, 28, 1294–1301. [CrossRef]
89. Stynski, S.; Luo, W.; Chub, A.; Franquelo, L.G.; Malinowski, M.; Vinnikov, D. Utility-Scale Energy Storage Systems: Converters
and Control. IEEE Ind. Electron. Mag. 2020, 14, 32–52. [CrossRef]
90. Inoue, S.; Akagi, H. A Bidirectional DC–DC Converter for an Energy Storage System with Galvanic Isolation. IEEE Trans. Power
91. Manias, S.; Ziogas, P.; Olivier, G. Bilateral DC to AC convertor using a high frequency link. IEE Proc. B Electr. Power Appl. 1987,
134, 15. [CrossRef]
92. Blinov, A.; Korkh, O.; Chub, A.; Vinnikov, D.; Peftitsis, D.; Norrga, S.; Galkin, I. High Gain DC-AC High-Frequency Link Inverter
with Improved Quasi-Resonant Modulation. IEEE Trans. Ind. Electron. 2021. [CrossRef]
93. Korkh, O.; Blinov, A.; Vinnikov, D.; Chub, A. Review of Isolated Matrix Inverters: Topologies, Modulation Methods and
Applications. Energies 2020, 13, 2394. [CrossRef]
94. Chung, H.; Cheung, W.-L.; Tang, K. A ZCS Bidirectional Flyback DC/DC Converter. IEEE Trans. Power Electron. 2004, 19,
1426–1434. [CrossRef]
95. Lessing, M.H.; Agostini, E.; Barbi, I. An improved modulation strategy for the high-frequency-isolated DC-AC flyback converter
with differential output connection. In Proceedings of the 2016 12th IEEE International Conference on Industry Applications
(INDUSCON), Curitiba, Brazil, 20–23 November 2016; Institute of Electrical and Electronics Engineers (IEEE): New York, NY,
USA, 2017; pp. 1–7.
96. Ponnaluri, S.; Linhofer, G.; Steinke, J.; Steimer, P.K. Comparison of single and two stage topologies for interface of BESS or fuel
cell system using the ABB standard power electronics building blocks. In Proceedings of the 2005 European Conference on Power
Electronics and Applications, Dresden, Germany, 11–14 September 2005; Institute of Electrical and Electronics Engineers (IEEE):
New York, NY, USA, 2005.
97. Srinivasan, R.; Oruganti, R. A unity power factor converter using half-bridge boost topology. IEEE Trans. Power Electron. 1998, 13,
98. Vazquez, S.; Lukic, S.M.; Galvan, E.; Franquelo, L.G.; Carrasco, J.M. Energy Storage Systems for Transport and Grid Applications.
IEEE Trans. Ind. Electron. 2010, 57, 3881–3895. [CrossRef]
99. Stanley, G.; Bradshaw, K. Precision DC-to-AC power conversion by optimization of the output current waveform-the half bridge
revisited. IEEE Trans. Power Electron. 1999, 14, 372–380. [CrossRef]
100. Zhang, L.; Zhu, T.; Chen, L.; Sun, K. A systematic topology generation method for dual-buck inverters. In 2016 IEEE Energy
Conversion Congress and Exposition (ECCE); Institute of Electrical and Electronics Engineers (IEEE): New York, NY, USA, 2017;
pp. 1–6.
101. Xie, J.; Zhang, F.; Ren, R.; Wang, X.; Wang, J. A novel high power density dual-buck inverter with coupled filter inductors. In
IECON 2014-40th Annual Conference of the IEEE Industrial Electronics Society; Institute of Electrical and Electronics Engineers (IEEE):
New York, NY, USA, 2014; pp. 1111–1117.
102. Zhou, L.; Gao, F. Dual buck inverter with series connected diodes and single inductor. In 2016 IEEE Applied Power Electronics
Conference and Exposition (APEC); Institute of Electrical and Electronics Engineers (IEEE): New York, NY, USA, 2016; pp. 2259–2263.
103. Araujo, S.V.; Zacharias, P.; Mallwitz, R. Highly Efficient Single-Phase Transformerless Inverters for Grid-Connected Photovoltaic
Systems. IEEE Trans. Ind. Electron. 2010, 57, 3118–3128. [CrossRef]
104. Gu, B.; Dominic, J.; Lai, J.-S.; Chen, C.-L.; Labella, T.; Chen, B. High Reliability and Efficiency Single-Phase Transformerless
Inverter for Grid-Connected Photovoltaic Systems. IEEE Trans. Power Electron. 2012, 28, 2235–2245. [CrossRef]
105. Colak, I.; Kabalci, E.; Bayindir, R. Review of multilevel voltage source inverter topologies and control schemes. Energy Convers.
Manag. 2011, 52, 1114–1128. [CrossRef]
106. Liu, C.; Cai, X.; Chen, Q. Self-Adaptation Control of Second-Life Battery Energy Storage System Based on Cascaded H-Bridge
Converter. IEEE J. Emerg. Sel. Top. Power Electron. 2018, 8, 1428–1441. [CrossRef]
107. Ling, Z.; Zhang, Z.; Li, Z.; Li, Y. State-of-charge balancing control of battery energy storage system based on cascaded H-
bridge multilevel inverter. In Proceedings of the 2016 IEEE 8th International Power Electronics and Motion Control Conference
(IPEMC-ECCE Asia); Institute of Electrical and Electronics Engineers (IEEE): New York, NY, USA, 2016; pp. 2310–2314.
108. Yoscovich, I.; Glovinsky, T.; Sella, G.; Galin, Y. SolarEdge Patent for HD-Wave Inverters-Distributed Power System Using Direct
Current Power Sources. U.S. Patent 9368964B2, 14 June 2016.
109. Demetriades, G.D. Modular Multilevel Converter with Cell-Connected Battery Storages. European Patent EP2695273B1,
25 November 2015.
110. Moradpour, M.; Ghani, P.; Pirino, P.; Gatto, G. A GaN-Based Battery Energy Storage System for Three-Phase Residential
Application with Series-Stacked Devices and Three-Level Neutral Point Clamped Topology. In 2019 1st International Conference on
Energy Transition in the Mediterranean Area (SyNERGY MED); Institute of Electrical and Electronics Engineers (IEEE): New York,
NY, USA, 2019; pp. 1–6.
111. Abronzini, U.; Attaianese, C.; Di Monaco, M.; Tomasso, G.; Damiano, A.; Porru, M.; Serpi, A. A Dual-Source DHB-NPC Power
Converter for Grid Connected Split Battery Energy Storage System. In 2018 IEEE Energy Conversion Congress and Exposition
(ECCE); Institute of Electrical and Electronics Engineers (IEEE): New York, NY, USA, 2018; pp. 2483–2488.
128
Energies 2021, 14, 3365
112. Trintis, I.; Teodorescu, R.; Munk-Nielsen, S. Single stage grid converters for battery energy storage. In Proceedings of the 5th IET
International Conference on Power Electronics, Machines and Drives (PEMD 2010), Brighton, UK, 19–21 April 2010; Institution of
Engineering and Technology (IET): London, UK, 2010; p. 234.
113. Bragard, M.; Soltau, N.; Thomas, S.; De Doncker, R.W. The Balance of Renewable Sources and User Demands in Grids: Power
Electronics for Modular Battery Energy Storage Systems. IEEE Trans. Power Electron. 2010, 25, 3049–3056. [CrossRef]
114. Cheng, Y.; Qian, C.; Crow, M.L.; Pekarek, S.; Atcitty, S. A Comparison of Diode-Clamped and Cascaded Multilevel Converters for
a STATCOM with Energy Storage. IEEE Trans. Ind. Electron. 2006, 53, 1512–1521. [CrossRef]
115. Barth, C.B.; Assem, P.; Foulkes, T.; Chung, W.H.; Modeer, T.; Lei, Y.; Pilawa-Podgurski, R.C.N. Design and Control of a GaN-Based,
13-Level, Flying Capacitor Multilevel Inverter. IEEE J. Emerg. Sel. Top. Power Electron. 2019, 8, 2179–2191. [CrossRef]
116. Modeer, T.; Pallo, N.; Foulkes, T.; Barth, C.B.; Pilawa-Podgurski, R.C.N. Design of a GaN-Based Interleaved Nine-Level Flying
Capacitor Multilevel Inverter for Electric Aircraft Applications. IEEE Trans. Power Electron. 2020, 35, 12153–12165. [CrossRef]
117. Siwakoti, Y.P.; Peng, F.Z.; Blaabjerg, F.; Loh, P.C.; Town, G.E. Impedance-Source Networks for Electric Power Conversion Part I: A
Topological Review. IEEE Trans. Power Electron. 2015, 30, 699–716. [CrossRef]
118. Mande, D.; Trovão, J.P.; Ta, M.C. Comprehensive Review on Main Topologies of Impedance Source Inverter Used in Electric
Vehicle Applications. World Electr. Veh. J. 2020, 11, 37. [CrossRef]
119. You, K.; Rahman, M.F. A Matrix–Z-Source Converter with AC–DC Bidirectional Power Flow for an Integrated Starter Alternator
System. IEEE Trans. Ind. Appl. 2009, 45, 239–248. [CrossRef]
120. Matiushkin, O.; Husev, O.; Strzelecki, R.; Ivanets, S.; Fesenko, A. Novel single-stage buck-boost inverter with unfolding circuit. In
Proceedings of the 2017 IEEE First Ukraine Conference on Electrical and Computer Engineering (UKRCON), Kyiv, Ukraine, 29
May–2 June 2017; Institute of Electrical and Electronics Engineers (IEEE): New York, NY, USA, 2017; pp. 538–543.
121. Han, B.; Lai, J.-S.; Kim, M. Bridgeless Cuk-Derived Single Power Conversion Inverter with Reactive-Power Capability. IEEE
Trans. Power Electron. 2019, 35, 2629–2645. [CrossRef]
122. Pal, A.; Basu, K. A Single-Stage Soft-Switched Isolated Three-Phase DC–AC Converter with Three-Phase Unfolder. IEEE Trans.
123. Tytelmaier, K.; Husev, O.; Veligorskyi, O.; Yershov, R. A review of non-isolated bidirectional dc-dc converters for energy storage
systems. In 2016 II International Young Scientists Forum on Applied Physics and Engineering (YSF); Institute of Electrical and
Electronics Engineers (IEEE): New York, NY, USA, 2016; pp. 22–28.
124. Anzola, J.; Aizpuru, I.; Romero, A.A.; Loiti, A.A.; Lopez-Erauskin, R.; Sevil, J.S.A.; Bernal, C. Review of Architectures Based on
Partial Power Processing for DC-DC Applications. IEEE Access 2020, 8, 103405–103418. [CrossRef]
125. Pape, M.; Kazerani, M. An Offshore Wind Farm with DC Collection System Featuring Differential Power Processing. IEEE Trans.
Energy Convers. 2019, 35, 222–236. [CrossRef]
126. Pape, M.; Kazerani, M. Turbine Startup and Shutdown in Wind Farms Featuring Partial Power Processing Converters. IEEE Open
Access J. Power Energy 2020, 7, 254–264. [CrossRef]
127. Iyer, V.M.; Gulur, S.; Gohil, G.; Bhattacharya, S. An Approach towards Extreme Fast Charging Station Power Delivery for Electric
Vehicles with Partial Power Processing. IEEE Trans. Ind. Electron. 2020, 67, 8076–8087. [CrossRef]
128. Schaef, C.; Stauth, J.T. Multilevel Power Point Tracking for Partial Power Processing Photovoltaic Converters. IEEE J. Emerg. Sel.
Top. Power Electron. 2014, 2, 859–869. [CrossRef]
129. Shenoy, P.S. Improving Performance, Efficiency, and Reliability of DC/DC Conversion Systems by Differential Power Processing.
Ph.D. Thesis, Graduate College of the University of Illinois at Urbana-Champaign, Champaign, IL, USA, 2012.
130. Zientarski, J.R.R.; Martins, M.L.D.S.; Pinheiro, J.R.; Hey, H.L. Evaluation of Power Processing in Series-Connected Partial-Power
Converters. IEEE J. Emerg. Sel. Top. Power Electron. 2019, 7, 343–352. [CrossRef]
131. Zientarski, J.R.R.; Martins, M.L.D.S.; Pinheiro, J.R.; Hey, H.L. Series-Connected Partial-Power Converters Applied to PV Systems:
A Design Approach Based on Step-Up/Down Voltage Regulation Range. IEEE Trans. Power Electron. 2018, 33, 7622–7633.
[CrossRef]
132. Zapata, J.W.; Kouro, S.; Carrasco, G.; Renaudineau, H.; Meynard, T.A. Analysis of Partial Power DC–DC Converters for Two-Stage
Photovoltaic Systems. IEEE J. Emerg. Sel. Top. Power Electron. 2019, 7, 591–603. [CrossRef]
133. Agamy, M.S.; Harfman-Todorovic, M.; Elasser, A.; Chi, S.; Steigerwald, R.L.; Sabate, J.A.; McCann, A.J.; Zhang, L.; Mueller, F.J.
An Efficient Partial Power Processing DC/DC Converter for Distributed PV Architectures. IEEE Trans. Power Electron. 2014, 29,
134. Marti-Arbona, E.; Mandal, D.; Bakkaloglu, B.; Kiaei, S. PV panel power optimization using sub-panel MPPT. In Proceedings of
the 2015 IEEE Applied Power Electronics Conference and Exposition (APEC), Charlotte, NC, USA, 15–19 March 2015; Institute of
Electrical and Electronics Engineers (IEEE): New York, NY, USA, 2015; pp. 235–238.
135. Loera-Palomo, R.; Morales-Saldaña, J.A.; Palacios-Hernández, E. Quadratic step-down dc–dc converters based on reduced
redundant power processing approach. IET Power Electron. 2013, 6, 136–145. [CrossRef]
136. Spiazzi, G. Reduced redundant power processing concept: A reexamination. In Proceedings of the 2016 IEEE 17th Workshop on
Control and Modeling for Power Electronics (COMPEL), Trondheim, Norway, 27–30 June 2016; pp. 1–8. [CrossRef]
137. Xue, F.; Yu, R.; Huang, A. Fractional converter for high efficiency high power battery energy storage system. In Proceedings of
the 2017 IEEE Energy Conversion Congress and Exposition (ECCE), Cincinnati, OH, USA, 1–5 October 2017; Institute of Electrical
and Electronics Engineers (IEEE): New York, NY, USA, 2017; pp. 5144–5150.
129
Energies 2021, 14, 3365
138. Xue, F.; Yu, R.; Huang, A. A Family of Ultrahigh Efficiency Fractional dc–dc Topologies for High Power Energy Storage Device.
IEEE J. Emerg. Sel. Top. Power Electron. 2021, 9, 1420–1427. [CrossRef]
139. Neumayr, D.; Knabben, G.C.; Varescon, E.; Bortis, D.; Kolar, J.W. Comparative Evaluation of a Full- and Partial-Power Processing
Active Power Buffer for Ultracompact Single-Phase DC/AC Converter Systems. IEEE J. Emerg. Sel. Top. Power Electron. 2021, 9,
1994–2013. [CrossRef]
140. Bala, S.; Tengner, T.; Rosenfeld, P.; Delince, F. The effect of low frequency current ripple on the performance of a Lithium Iron
Phosphate (LFP) battery energy storage system. In 2012 IEEE Energy Conversion Congress and Exposition (ECCE); Institute of
Electrical and Electronics Engineers (IEEE): New York, NY, USA, 2012; pp. 3485–3492.
141. Brand, M.J.; Hofmann, M.H.; Schuster, S.S.; Keil, P.; Jossen, A. The Influence of Current Ripples on the Lifetime of Lithium-Ion
Batteries. IEEE Trans. Veh. Technol. 2018, 67, 10438–10445. [CrossRef]
142. Husev, O.; Matiushkin, O.; Roncero-Clemente, C.; Blaabjerg, F.; Vinnikov, D. Novel Family of Single-Stage Buck–Boost Inverters
Based on Unfolding Circuit. IEEE Trans. Power Electron. 2018, 34, 7662–7676. [CrossRef]
143. Zhao, B.; Abramovitz, A.; Liu, C.; Yang, Y.; Huangfu, Y. A Family of Single-Stage, Buck-Boost Inverters for Photovoltaic
Applications. Energies 2020, 13, 1675. [CrossRef]
144. Oguchi, K.; Ikawa, E.; Tsukiori, Y. A three-phase sine wave inverter system using multiple phase-shifted single-phase resonant
inverters. IEEE Trans. Ind. Appl. 1993, 29, 1076–1083. [CrossRef]
145. Rabkowski, J.; Blinov, A.; Zinchenko, D.; Wrona, G.; Zdanowski, M. Grid-frequency Vienna rectifier and isolated current-source
DC-DC converters for efficient off-board charging of electric vehicles. In Proceedings of the 2020 22nd European Conference on
Power Electronics and Applications (EPE’20 ECCE Europe), Piscataway, NJ, USA, 7–11 September 2020; IEEE: New York, NY,
USA, 2020.
146. Jacobson, B.S.; Holmansky, E.N. Methods and Apparatus for Three-Phase Inverter with Reduced Energy Storage. U.S. Patent
7839023B2, 23 November 2010.
147. Chen, W.W.; Zane, R.; Corradini, L. Isolated Bidirectional Grid-Tied Three-Phase AC–DC Power Conversion Using Series-
Resonant Converter Modules and a Three-Phase Unfolder. IEEE Trans. Power Electron. 2017, 32, 9001–9012. [CrossRef]
148. Van Soest, H. Peer-to-peer electricity trading: A review of the legal context. Compet. Regul. Netw. Ind. 2018, 19, 180–199. [CrossRef]
149. Zhang, D.; Guacci, M.; Kolar, J.W.; Everts, J. Synergetic Control of a 3-Φ Buck-Boost Current DC-Link EV Charger Considering
Wide Output Range and Irregular Mains Conditions. In Proceedings of the 2020 IEEE 9th International Power Electronics and
Motion Control Conference (IPEMC2020-ECCE Asia), Nanjing, China, 31 May–3 June 2020; pp. 1688–1695.
150. Husev, O.; Roncero-Clemente, C.; Romero-Cadaval, E.; Vinnikov, D.; Stepenko, S. Single phase three-level neutral-point-clamped
quasi-Z-source inverter. IET Power Electron. 2015, 8, 1–10. [CrossRef]
151. Panfilov, D.; Husev, O.; Blaabjerg, F.; Zakis, J.; Khandakji, K. Comparison of three-phase three-level voltage source inverter with
intermediate dc–dc boost converter and quasi-Z-source inverter. IET Power Electron. 2016, 9, 1238–1248. [CrossRef]
152. Baier, T.; Piepenbreier, B. Bidirectional magnetically coupled T-Source Inverter for extra low voltage application. In Proceedings
of the 2016 IEEE Applied Power Electronics Conference and Exposition (APEC), Long Beach, CA, USA, 20–24 March 2016; pp.
2897–2904.
153. Hu, S.; Liang, Z.; He, X. Ultracapacitor-Battery Hybrid Energy Storage System Based on the Asymmetric Bidirectional Z -Source
Topology for EV. IEEE Trans. Power Electron. 2016, 31, 7489–7498. [CrossRef]
130
energies
Article
Design of an Effective State of Charge Estimation Method for a
Lithium-Ion Battery Pack Using Extended Kalman Filter and
Artificial Neural Network
Van Quan Dao 1 , Minh-Chau Dinh 2 , Chang Soon Kim 2 , Minwon Park 1 , Chil-Hoon Doh 3 , Jeong Hyo Bae 3 ,
Myung-Kwan Lee 4 , Jianyong Liu 5 and Zhiguo Bai 5, *
1 Department of Electrical Engineering, Changwon National University, Changwon 51140, Korea;

quandao.hust98@gmail.com (V.Q.D.); paku@changwon.ac.kr (M.P.)
2 Institute of Mechatronics, Changwon National University, Changwon 51140, Korea;
thanchau7787@gmail.com (M.-C.D.); ee.cskim@gmail.com (C.S.K.)
3 Distributed Power System Research Center, Korea Electrotechnology Research Institute,
Changwon 51543, Korea; chdoh@keri.re.kr (C.-H.D.); jhbae@keri.re.kr (J.H.B.)
4 Battery Solution Co., Ltd., Jeonnam 58324, Korea; kem4328@naver.com
5 IES Co., Ltd., Busan 46744, Korea; newsir86@gmail.com
* Correspondence: kawabai@gmail.com; Tel.: +82-10-8310-2595
Abstract: Currently, Lithium-ion batteries (LiB) are widely applied in energy storage devices in smart
grids and electric vehicles. The state of charge (SOC) is an indication of the available battery capacity,
and is one of the most important factors that should be monitored to optimize LiB’s performance
and improve its lifetime. However, because the SOC relies on many nonlinear factors, it is difficult
to estimate accurately. This paper presented the design of an effective SOC estimation method for
Citation: Dao, V.Q.; Dinh, M.-C.;
a LiB pack Battery Management System (BMS) based on Kalman Filter (KF) and Artificial Neural
Kim, C.S.; Park, M.; Doh, C.-H.; Bae,
Network (ANN). First, considering the configuration and specifications of the BMS and LiB pack,
J.H.; Lee, M.-K.; Liu, J.; Bai, Z. Design
an ANN was constructed for the SOC estimation, and then the ANN was trained and tested using
of an Effective State of Charge
Estimation Method for a Lithium-Ion
the Google TensorFlow open-source library. An SOC estimation model based on the extended
Battery Pack Using Extended Kalman KF (EKF) and a Thevenin battery model was developed. Then, we proposed a combined mode
Filter and Artificial Neural Network. EKF-ANN that integrates the estimation of the EKF into the ANN. Both methods were evaluated
Energies 2021, 14, 2634. https:// through experiments conducted on a real LiB pack. As a result, the ANN and KF methods showed
doi.org/10.3390/en14092634 maximum errors of 2.6% and 2.8%, but the EKF-ANN method showed better performance with less
than 1% error.
Academic Editor: Andrei Blinov
Keywords: Artificial neural network; battery management system; Kalman filter; lithium-ion battery;
Received: 6 April 2021
state of charge estimation
Accepted: 27 April 2021
Published: 4 May 2021

1. Introduction
Energy storage systems are emerging as the biggest concern for modern smart grids
iations. and electric vehicles (EV), and the lithium-ion battery (LiB) technology is an efficient
solution for energy storage applications with the advantages of long cycle life, large
capacity and no memory effect. Already commercialized and matured for consumer
electronic applications, the LiB is being positioning itself as a leading technology platform
for plug-in hybrid electric vehicles (PHEVs) and all EVs [1,2]. It is also widely used in
large facilities to support energy storage [3], load-leveling and peak shaving in the power
grid [4], frequency regulation [5], and to reduce network load and capacity payments [6] in
the smart grids.
conditions of the Creative Commons In order for the LiBs to work as expected, a battery management system (BMS) must
Attribution (CC BY) license (https:// be designed for tracking and controlling the current level of battery energy. The BMS is
creativecommons.org/licenses/by/ defined as an electronic equipment that manages a rechargeable battery (single cell or
4.0/). battery pack). The main functions of the BMS are to monitor, compute, communicate,

Energies 2021, 14, 2634
protect, and optimize [7]. At this point, the estimate of state of charge (SOC) is one of
the critical functions in the BMS. The SOC is defined as the percentage of the available
capacity to the rated capacity of the battery, and many issues with the LiB, such as capacity
degradation, increased maintenance costs, rapid aging, serious equipment failures, and
even dangerous accidents, are related to incorrect SOC estimates [8]. Therefore, an accurate
estimation of the SOC is very important for optimizing battery performance, including
extending battery life and preventing permanent damage to the batteries.
In general, the battery SOC nonlinearly depends on several factors including current,
voltage, temperature, and battery aging [9]. Therefore, an accurate estimate of the SOC
is quite complicated. Various techniques have been presented to estimate the SOC of a
battery cell or a battery pack. Key technologies include discharge tests, open-circuit voltage
measurement, Coulomb counting, inherent resistance measurement, and intelligent SOC
estimation methods [10,11]. Intelligent computation techniques such as artificial neural
network (ANN) and Kalman filter (KF) have been developed for EV applications [12–16].
Compared to other techniques, they have several advantages such as high accuracy, real-
time calculation, simple current and voltage measurements. Specifically, these techniques
are highly adaptable to the dynamic behaviors of batteries due to their self-learning ability.
However, there are still some issues that need to be studied. Applying KF requires accurate
battery modeling, and important factors such as temperature and SOC that may affect the
internal parameters of the battery model are not yet considered. Using an ANN requires
a large amount of training data that can lead to a large dimension and high computa-
tion of the network when implemented in a BMS. Therefore, it is necessary to design a
practical BMS to properly analyze and evaluate the operational characteristics of the SOC
estimation methods.
In this paper, an effective SOC estimation method was designed and implemented in
a smart BMS for a LiB pack based on the extended KF (EKF) and ANN. First, the structure
and specifications of the smart BMS and LiB pack were summarized, and the design
process of the ANN was described in detail. The ANN was then trained and tested for SOC
estimation using real battery data sets. Next, we developed an SOC estimation algorithm
based on the EKF and a Thevenin battery model. Finally, we proposed a combination
model of EKF and ANN (EKF-ANN) to compensate for the shortcomings of the above
two methods. To evaluate the effectiveness of the SOC estimation method, the proposed
methods were experimentally verified and compared with each other. As a result, the
proposed ANN and EKF methods showed an error of 2.6% and 2.8%, respectively, and the
SOC estimation error when using the EKF-ANN was significantly improved to less than
1%. The results show that the proposed SOC estimation method satisfies the requirements
of the BMS for LiB packs.
2. Review of SOC Estimation Methods

An accurate estimate of the SOC plays an important role in a credible BMS, but the
SOC cannot be measured directly. The SOC is associated with direct measurements such as
current, voltage, temperature, and it can be extracted based on intrinsic relations or control
theory of the battery. Many techniques have been proposed to estimate the battery SOC. In
this section, we discuss some popular SOC estimation methods and compare them with
each other.
2.1. Open Circuit Voltage Method

This method calculates the SOC or the remaining capacity of the battery based on the
measured open-circuit voltage (OCV). Each battery has a corresponding curve between
the SOC and OCV, from which one can determine the other. However, in order to get
a stable voltage, the battery must be rested for a long time under no load. Moreover,
since the OCV-SOC curve is sensitive to various temperatures and discharge rates, the
method is only effective in estimating the SOC at the early and end stages of the charging
132
Energies 2021, 14, 2634
and discharging process after the battery has been disconnected from the load for a long
time [17,18].
2.2. Coulomb Counting Method

Coulomb counting, which is also called Ampare-hour (Ah) counting, is the most
common technique for estimating the SOC based on the integration on time of the charge
and discharge current values. However, the initial SOC is difficult to determine at the
starting state. Even though we can gain the SOC from the record of the BMS or OCV
look-up table, but the accuracy is hard to ensure [19]. Additionally, the SOC calculation is
based only on the measurement of current without considering the measurement noise.
Over time, errors will be accumulated due to the integration factor, and this is the reason
why this method is prone to errors.
2.3. Impedance-Based Method

There is a dependency between the SOC and the impedance of a battery, and thus
the SOC can be considered a function of battery impedance change [20]. However, the
impedance varies significantly with the aging status of a battery; thus, this technology is no
longer a good indicator for the SOC. Furthermore, the sensitivity of the battery impedance
on the temperature is very high; thereby a high accuracy of SOC estimation is impossible to
maintain for batteries in EVs due to quick temperature change during the driving process.
2.4. Kalman Filter-Based Method

The KF is a method for determining the internal states of any dynamic process,
in which the mean of the squared error is minimized. Its target is to obtain accurate
information from inaccurate data. This method can be utilized to calculate the SOC in
real-time by using the terminal current and voltage measurements [21–23]. It is suitable for
the SOC estimation of EVs in which the battery current is unstable [24]. However, it has
high demands for the battery modelling and computational capability [25,26].
2.5. Artificial Neural Network-Based Method

The ANN is an intelligent technology, which has a strong self-learning and high
adaptability, and this technique is very useful for researching complex nonlinear system
models. For the SOC estimation, the ANN is able to be applied in all battery systems
without the information of cell internal structure, as long as the battery dataset for training
the network is available [27,28]. Also, the ANN has the ability to estimate the SOC without
the initial SOC.
2.6. Fuzzy Logic-Based Method

This method is based on simulating the fuzzy thinking of a human being using
the fuzzy logic based on a large number of test curves, experience and reliable fuzzy
logic theories, and finally to perform the SOC estimation [29]. It requires a complete
understanding of the battery itself and relatively large computations. However, the battery
parameters significantly vary with the battery lifetime, and so the SOC estimation may
not be accurate enough. It is only suitable for static battery characteristics and practically
inapplicable for LiBs in the EVs [30].
In this paper, we decided to develop an effective SOC estimation method based on the
ANN and KF.
3. Design of SOC Estimation Methods for a Smart BMS

3.1. Configuration of the Smart BMS and LiB Pack
A BMS is a device including hardware and software which controls the operating
conditions of the battery to extend battery life, ensure safety, and accurately predict the
various states of the battery. To ensure this, the BMS has several functions to control and
monitor the battery states at various battery cells, battery modules, and battery pack levels.
133
Energies 2021, 14, 2634
The estimation of the SOC is a core function in a BMS, but it is still a challenge to accurately
estimate in online-real time, as battery characteristics change with the degree of aging and
strong nonlinear and complex electrochemical reactions of the battery. The SOC can be
estimated by continuously measuring terminal voltage, current and temperature of the
battery. For a battery system in EVs and smart grids, the current trend is the design of
a smart BMS, which includes researches in the field of artificial intelligence (AI) utilized
for the battery SOC estimation. In large battery pack applications, the predictability
and adaptability of BMS are especially important. In this study, in order to verify the
performance of the SOC estimation methods through experiments, we developed a smart
BMS module based on the master and slave topology design for the LiB pack as shown in
Figure 1.
Figure 1. Configurations of the smart BMS and LiB pack.
The LiB pack consisted of 3 module units connected in series, and each module unit
consisted of 36 cells with a 9S4P configuration, meaning 9 cells in series and 4 cells in
parallel. There are 3 slave microcontroller unit (MCU) boards to manage the module
units. The main MCU board, which is the master, coordinates two-way communication
between the master and the slave and executes battery management actions. The detailed
specifications of the smart BMS and LiB pack are described in Table 1. We will develop
SOC estimation methods with an error target of 3% as shown in Table 2.
Table 1. Detailed specifications of the smart BMS and LiB pack.
Items Specifications Contents

Master BMS MCU 32-bit STM32F207
Slave BMS MCU 32-bit STM32F103
External communication UART
1. BMS hardware Master-Slave communication CAN 2.0
Power supply 12 V
Temperature sensors 4 ea.
Firmware tool ST-Link Download
Model 18,650 35E
2. LiB cell Capacity 3.4 Ah (1 C)
Cell voltage 2.5~4.2 V
134
Energies 2021, 14, 2634
Table 1. Cont.
Items Specifications Contents

Configuration 9-series and 4-parallel (9S4P)
3. LiB unit module Voltage/Capacity 32.4 V/10.2 Ah
Energy 330.48 Wh
Configuration 3 LiB units in series
Pack voltage 67.5~113.4 V
4. LiB pack
Capacity 10.2 Ah
Energy 991.44 Wh
Table 2. Design targets of the smart BMS.
Items Standard [31,32] Target Unit Evaluation Method

Active Cell Balancing ±10 ±8 mV Digital multimeter
Voltage measurement accuracy ±5 ±1 mV Digital multimeter
Current measurement accuracy ±2 ±1 % Charger/Discharger
Temperature measurement accuracy ±1 ±0.5 ◦C Thermostat
SOC estimation error ±5 ±3 % Charger/Discharger
3.2. SOC Estimation Method Using an ANN

3.2.1. Design Process for the ANN
An ANN is a computational model constructed by a set of individual processing units,
which are called the artificial neurons. These neurons are interconnected by weights. The
ANN is a universal approximator that is able to model any nonlinear function with the
desired precision [33]. The network is configured in layers, with the input layer receiving
inputs and the output layer generating outputs. The middle layers, called the hidden
layers, have no connection with the outside. The neurons are responsible for connecting
each layer, thereby they are the central component of the ANN. The basic configuration of
a neuron in the network is described in Figure 2. The vector of the input signals is denoted
by X = [ x1 , x2 , x3 , . . . , xn ], n ∈ N, the neuron weights by W = [W1 , W2 , W3 , . . . , Wn ], net is
the multiplication response of weights with the input signals, b1 is an external parameter
called as the bias, f is the activation function, and ym is the output signal of the neuron.
Figure 2. Configuration of a neuron in the ANN.
Table 3 shows the design process and design factors of an ANN for the SOC estimation.
It starts from collecting and pre-processing the data, then designing the neural network in
terms of network topology, configuration, activation function, loss function and metrics,
loss optimizer, and learning rate. Finally, the training and validation of the network are
implemented.
135
Energies 2021, 14, 2634
Table 3. Design process and design factors for the ANN.
Step Contents Remarks
- Training dataset (%)

Data collection and pre-processing - Validation dataset (%)
1
for training the network - Test dataset (%)
- Input layer: set the inputs

Selecting the network topology and - Number of hidden layers, and number
2 of neurons in each layer
configuring the layers and neurons
- Output layer: set the output
3 Selecting the learning rate, α - 0<α<1
4 Selecting the activation function - Sigmoid, Linear, Tanh, ReLu, . . .
- Mean squared error (MSE)

Selecting the loss function and - Mean absolute error (MAE)
5
metrics for the training process - Metrics = [MSE, MAE]
Selecting the learning function - Adam, AdaDelta, SGD, . . .
6
(optimizer)
3.2.2. Design of the ANN for the SOC Estimation

The historical battery dataset was collected from the designed LiB pack by the com-
pany Battery Solution Co., Ltd. in South Korea. The dataset consists of the test time in
seconds, the pack current in ampere, the pack terminal voltage in volts, the pack tempera-
ture, and the cumulative charge/discharge capacity of the battery pack in ampere-hour.
The LiB pack was tested under 0–100% of the SOC condition for 20 cycles at room tempera-
ture. The data was collected with a sampling time of 1 s. The battery pack was charged and
discharged with a constant current of 1.7 A (0.5 C rate). The battery charge and discharge
processes correspond to the plus (+) and minus (−) signs of the current. The temperature
was measured at the center of the battery pack by a thermal sensor with an accuracy of
±0.5 ◦ C, and the temperature range was from 19 ◦ C to 42 ◦ C. From the battery pack dataset,
the average values of voltage and current in one cell were calculated. The cell voltage
range was from 2.6 V to 4.2 V. The actual SOC was also calculated offline by the integration
on time of current to use for training and testing the ANN model. The dataset must be
processed to obtain a satisfactory ANN, all trash data was deleted. The ANN was trained
to estimate the SOC of one cell.
The next step was to design the network topology for the ANN, and the Multi-Layer
Perception (MLP) network was selected. The inputs of the network were the current,
voltage, and temperature, and the output was the battery SOC. Figure 3 describes the
relationships between the inputs and outputs of the designed ANN. The SOC has strong
and nonlinear relationships with the voltage and temperature, which can be expressed
as exponential functions. Thus, the nonlinear activation functions including hyperbolic
tangent (tanh) and logistic (sigmoid) were considered for the ANN. The learning rate, α, is
an amount by which the weights are updated during the training process. In the proposed
network, the learning rate was set at 0.001.
136
Energies 2021, 14, 2634
Figure 3. Relationships between the inputs and output of the ANN.
After tuning the parameters and testing several configurations of the network, the
most suitable specifications for the ANN are described in Table 4.
Table 4. Final specifications of the ANN for the SOC estimation.
Specifications ANN
Number of inputs of the network 3
Number of hidden layers of the network 4
Number of outputs of the network 1
Number of neurons in each hidden layer 64
Activation function in hidden layers Sigmoid
Activation function in the output layer Softplus
Adapting learning function Adam
Learning rate 0.001
Loss function MSE
Accuracy metrics [MSE, MAE]
The battery dataset contained approximately 250,000 samples, corresponding to the

data measured at each sampling time. Each sample consisted of three input variables
(battery voltage, current and temperature) and the actual SOC. Before training the network,
the dataset was normalized and randomized to separate them into three sets: training,
validation, and testing. The data in each set were different from each other and split
evenly into three input variables. The amount of sample data for the training, validation,
and testing were 80%, 10%, and 10% of the total samples, respectively. In this study, the
ANN was designed, trained, and tested by using the Keras framework in the TensorFlow
open-source library of Google. The ANN was trained during 500 epochs. As a result, the
training performance achieved an MSE of approximately 0.5 and an MAE of approximately
0.43 in the final epoch as shown in Figure 4. From the epoch of 150, the training losses
almost match the validating losses.
(a) (b)
Figure 4. Training performances of the ANN: (a) Mean squared error (MSE); and (b) Mean absolute error (MAE).
137
Energies 2021, 14, 2634
After the training process, the model testing was performed to compare the estimated
results and the actual results as described in Figure 5. The trained ANN was tested
randomly with 5,000 samples, and the maximum and average errors of the estimated SOC
were 2.3% and 0.34%, respectively. This result is satisfied with the target of SOC estimation.
(a) (b)
Figure 5. Test results of the trained ANN with 5,000 random samples: (a) error of the estimated SOC, and (b) estimated and
actual SOCs.
When the training process was completed, the weights and bias of the network were
saved and exported as matrices to implement the SOC estimation function in the BMS for
experiments.
3.3. SOC Estimation Method Using the EKF

In 1960, Kalman successfully introduced the state space into filter theory and first
proposed the KF. Basically, the standard KF is only appropriate for linear dynamic systems
to obtain the best and unbiased estimate of the state variables. However, the battery is a
non-linear system; thus, it needs to be linearized and approximated to a linear system. At
this point, the system states can be estimated by the KF. This method is called the extended
Kalman Filter (EKF) [34]. To implement the EKF, a battery model is first developed to define
internal state variables, state-space equations to build the mathematical model. Moreover,
based on the external variables, such as terminal voltage, current, and temperature, the
battery model is able to determine the internal state variables such as the internal resistance,
electromotive force, capacitance and SOC.
3.3.1. Battery Model for the LiB

In this study, the Thevenin model was chosen as a LiB model. This model type has
good performance and can accurately simulate the dynamic characteristics of the LiB [35].
Figure 6 describes the equivalent circuit of the Thevenin battery model, which consists of
an ideal source of voltage VOC (OCV), an internal resistance R0 , a polarization resistance
Rp, and a capacitor Cp . The terminal voltage and current of the battery are denoted by VL
and IL , respectively. The voltage and current flowing through the branch Rp Cp are denoted
by Vp and Ip , respectively.
138
Energies 2021, 14, 2634
Figure 6. Equivalent circuit of the Thevenin model for the LiB.
Based on the Kirchhoff’s laws, the capacitor voltage variation, its current relationship
and the characteristics of the Thevenin model can be expressed as follows:
· IL VP
VP = − (1)
CP R P CP
VL = VOC − VP − IL R0 (2)
From Equations (1) and (2), where τ = R p C p is the time constant and Δt is the sam-
pling time, a discrete-time system for the VL and Vp is built as shown in the
Equations (3) and (4).
−Δt −Δt
VP (k ) = exp( ) × VP (k − 1) + (1 − exp( )) × IL (k − 1) R P (k) (3)
τ τ
VL (k ) = VOC (k ) − VP (k) − IL (k ) R0 (k ) (4)
The VOC , R0 , Rp and Cp are nonlinear functions of the battery SOC [36], thus the
Equations (3) and (4) can also be expressed as below.
−Δt −Δt
VP (k) = exp( ) × VP (k − 1) + (1 − exp( )) × IL (k − 1) R P (SOC (k)) (5)
τ τ
VL (k ) = VOC (SOC (k )) − VP (k) − IL (k ) R0 (SOC (k)) (6)
The SOC describes the relationship between the remaining and the maximum capacity
available in the battery and can be described as Equation (7):
IL (k ) × Δt
SOC (k ) = SOC (k − 1) − (7)
Ca
where Ca is the nominal capacity of the battery, and the SOC(k − 1) is the priority SOC.
By using the Equations (5)–(7), we can express the state equation for the nonlinear battery
system as follows:
xk = Ak−1 xk−1 + Bk−1 uk−1 + wk−1
(8)
yk = Ck xk + Dk uk + vk

VP (k ) 1 0 RP exp −τΔt
with, xk = , Ak = , Bk = ,
SOC (k ) 0 exp −τΔt Δt
Ca
Ck = dSOCdVoc
− dSOC
dRo
−1 , Dk = [− R0 (k)].
Where xk and uk are the system state vectors and the measured system inputs at
discrete-time k, and wk is the unmeasured process noise that can influence the system
state. The output of the system is yk , and vk is the measurement noise. Ak , Ck are the first
partial derivatives matrices with respect to xk , and Bk , and Dk are the Jacobian matrice
with respect to uk .
With the matrice Ck , dSOCdVoc
− dSOC
dRo
= Ms is a function of the SOC, and thus it can be
built from the battery experiment data.
139
Energies 2021, 14, 2634
3.3.2. Parameter Identification

In this section, we identify the internal parameters of the built battery model including
VOC , R0 , Rp , Cp, and Ms . This requires the experiment data including current and voltage
across battery terminals. The VOC of the battery is the potential difference between the poles
of the battery in the no load state. In order to obtain open-circuit voltage and the battery
SOC curve, the battery was charged in the standard way and then completely discharged
with a constant current. The battery was discharged from 100% to 0% in 5% steps, and the
period between the two steps is in a no-load (open-circuit) state; this procedure is completed
within 2 h to achieve stable VOC . From obtained data, the mathematical relationship
between the SOC and VOC was built by using the nonlinear curve fitting, and the fitting
results are shown in Equation (9) and Figure 7.
VOC = 2.926 + 0.044 × SOC1 − 0.0012 × SOC2 + 1.511E−5 × SOC3 − 6.72E−8 × SOC4 (9)
Figure 7. VOC and SOC curve of the LiB.
In the case of R0 , it can also be determined by using the above battery discharge data.
The value of R0 was calculated corresponding to the SOC of each period by dividing the
value of the voltage drop by the discharge current, where voltage drop is the difference
between the no load and loaded voltages. The variation of R0 according to the SOC is given
in Figure 8, which fits into Equation (10).
R0 = 5.7E−3 − 3.4E−2 SOC1 + 0.146SOC2 − 0.326SOC3 + 0.415SOC4 − 0.29SOC5 + 0.094SOC6 − 6.4E−3 SOC7 (10)
Figure 8. R0 and SOC curve of the LiB.
140
Energies 2021, 14, 2634
From the results of VOC and R0 , we can define the functions of Ms according to the
SOC as below.
Ms = −14.885+130.43 × SOC1 − 588.464 × SOC2 + 1381.381 × SOC3 − 1889.537 × SOC4 (11)

In order to identify the remaining battery parameters, Rp and Cp , the Matlab toolbox
“Parameter estimate”, which is a powerful interface for battery parameter identification,
was used. The obtained functions of Rp and Cp according to the SOC are shown in Equations
(12) and (13), respectively. Figure 9 describes the variation of Rp and Cp according to
the SOC.
R P = 0.015 − 0.18SOC1 + 1.052SOC2 − 3.28SOC3 + 5.798SOC4 − 5.82SOC5 + 3.08SOC6 − 0.67SOC7 (12)

CP = 1.3E3 − 2.8E4 × SOC + 2.6E5 × SOC2 − 1.09E6 × SOC3 + 2.4E6 × SOC4 − 3.08E6 × SOC5 + 2.01E6 × SOC6 − 5.3E5 × SOC7 (13)
(a) (b)
Figure 9. Variation of Rp and Cp according to SOC: (a) Rp; and (b) Cp.
3.3.3. SOC Estimation Model with the EKF

This algorithm combines the Thevenin battery model with the Ah integration method;
the SOC and Vp represent the system state variables, the current represents input variables,
and the battery terminal voltage represents the output variable. Figure 10 shows the
diagram of the SOC estimation method using the EKF.
Figure 10. Diagram of the SOC estimation method based on the EKF.
The use of the EKF requires the linearization of the state equations around the operat-
ing point for each sample. This algorithm is described in detail as follows:
141
Energies 2021, 14, 2634
- Definition of the battery model with the state Equation (8)

- Definition of the noise matrices: Q = E vk , vkT and R = E wk , wkT
- Initializing the state variable, x0+ = x0 and the estimation error P0+ = P0
- Prediction of the state variables, xk = [SOC (k ), Vp (k )] at time k by the Equations (6)
and (7) from the measured current IL (k )

∂ f ( xk ,uk ) ∂h( xk ,uk )
- Linearization by calculating Jacobians: F = ∂xk + and H = ∂x +
xk k xk
+
- Prediction of the covariance matrix Pk (estimation error): Pk = F · Pk−1 · F + Q T
−1
- Computation of the Kalman gain, Kg: K g = Pk · H T · ( H · Pk · H T + R)
- Update the new SOC based on the measured voltage, VL (k ) and the output variable
VL (k):

SOC (k )+ = SOC (k ) + K g · V (k ) − V (k) (14)
- Update the estimation error, Pk to Pk+ for the next time step (k + 1): Pk+ = ( IL (k) − K g ·
H ) · Pk
3.4. Combination of the EKF and ANN for the SOC Estimation
In the case of using the EKF, the battery model was built for the SOC estimation based
on the voltage and current characteristics. However, the battery is a dynamic system, and
with a long operation time, the internal parameters of the battery model can be changed
by other factors such as temperature, load changes, and SOC. These can cause errors in
the SOC estimation model, and the battery modeling is very complex when considering
them. With the method using the ANN, there were only three inputs including the voltage,
current, and temperature of the present sample, and there was no information in the
previous sample. This can cause a large dimension and high computation on the network.
As the above results, the designed network had 4 hidden layers with 64 neurons in each
layer, which means that the weight and bias matrices can take over a large amount of
memory in the MCU and make it difficult to handle computation.
To solve the above-mentioned problems, we suggested a SOC estimation method
combining the EKF with ANN as shown in Figure 11. The EKF model was applied to
determine the battery SOC in the previous sample, SOC(k − 1), based on the measured
values of the previous voltage, VL (k − 1) and the current, IL (k − 1). This SOC value was
used as one input of an ANN to estimate the SOC in the present sample, SOC(k). The ANN
was designed with four inputs including the measured voltage, VL (k), current, IL (k), and
temperature, T(k) at the present sample. The EKF is incorporated with the ANN to adapt
to the dynamic environments, and the SOC(k − 1) generated by the EKF takes into account
battery hysteresis and measurement noise. Thereby, the accuracy and reliability of the SOC
estimation can be improved.
Figure 11. Configuration of the SOC estimation method combining the EKF with ANN.
In this method, the designed EKF model in the previous section was kept to use,
and the ANN was designed and trained again with the same battery dataset. The final
specifications of the ANN are shown in Table 5. This network only needs two hidden
layers, and each hidden layer had 32 neurons. The required memory and computation time
142
Energies 2021, 14, 2634
in the MCU are significantly reduced. The detailed configuration of the designed ANN
including the input layer, hidden layers, output layer, number of neurons in each layer,
weight matrices, bias matrices, and activation functions are described in Figure 12.
Table 5. Final specifications of the ANN combined with the EKF.
Specifications ANN
Number of inputs of the network 4
Number of hidden layers of the network 2
Number of outputs of the network 1
Number of neurons in each hidden layer 32
Activation function in the hidden layers Sigmoid
Activation function in the output layer Softplus
Adapting learning function Adam
Learning rate 0.001
Loss function MSE
Accuracy metrics [MSE, MAE]
Figure 12. Detailed configuration and parameters of the designed ANN.
The network training was also performed in 500 epochs, and the performances of
the redesigned ANN are shown in Figure 13. The MSE and MAE in the final epoch are
approximately 0.0064 and 0.057, respectively, which are much lower than that of the ANN
using three inputs. The trained model was also tested randomly with 5,000 samples to
compare the actual and estimated SOC. The maximum and average errors of the estimated
SOC were 0.27% and 0.069%, respectively, as shown in Figure 14.
(a) (b)
Figure 13. Training performances of the redesigned ANN: (a) MSE, and (b) MAE.
143
Energies 2021, 14, 2634
(a) (b)
Figure 14. Test results of the redesigned ANN with 5,000 random samples: (a) error of estimated SOC, and (b) estimated
and actual SOCs.
4. Experiment Results and Discussions

4.1. Implementations of the SOC Estimation Methods in the Smart BMS
To validate the proposed SOC estimation methods, we applied them to the SOC
estimation function of the BMS in an experimental test with a real LiB pack. First, we
implemented a real-time SOC estimation based on the ANN. Then, the SOC estimations
based on the EKF and the EKF-ANN were implemented in offline experiments by using the
MATLAB program. These models were built to calculate the SOC from the experiment data
acquired using the ANN. Finally, the obtained results of three methods were compared
with the reference SOC, which was calculated from the measured battery capacity.
Using the ANN, the SOC was estimated based on the average voltage and current
of the battery cells and the temperature in the battery pack. The SOC estimation function
was built by C programming language on the master MCU STM32F205, which has a flash
memory of up to 1 Mbyte. In the designed ANN, there were three 64 × 64 weight matrices
and four 1x64 bias matrices in the hidden layers, and it took about 24 ms for calculation
speed. The SOC was calculated in each sampling time of 1 s with the full charge and
discharge of the battery pack. We found no problems with the memory and operation of
the master MCU during the experiment. A monitoring system was also facilitated for the
BMS as shown in Figure 15. The main interface of the monitoring system includes the
contents listed below:
(1) Configuration of the BMS
(2) Measurement values including SOC, local time, and BMS version
(3) Status of the BMS including warming, detection, state, and status
(4) Summary of main measurement values
(5) Serial communications with the BMS
(6) Command to read and write the alarm and cut-off values
(7) Calibration of the pack voltage and current values
(8) External inputs to block the relay
(9) Main screen displaying measurement values in the master and slave MCUs
144
Energies 2021, 14, 2634
Figure 15. Monitoring system for the smart BMS.
4.2. Experiment Results of the SOC Estimation Methods

Since the proposed SOC estimation methods were built based on the battery data,
these methods only valid under the experimental conditions similar to the range of battery
data collected as follows:
- After measuring the voltage and current of the entire battery pack, calculate the
average value of the voltage and current of one cell,
- Battery pack is discharged and charged with a constant current of 1.7 A in each cell,
- Cell voltage range is 2.6–4.2 V,
- Temperature range of the battery pack is 19~42 ◦ C.
In the experiment, we first performed the SOC estimation using the ANN in real-time.
The average voltage and current of the cells and the measured temperature during the
discharge and charge processes of the battery pack are described in Figure 16. The battery
pack was fully discharged and charged with the same currents of 1.7 A in each cell, which
were similar to the trained battery dataset. The temperature was measured at the center of
the battery pack, and the temperature range was from 19 ◦ C to 38 ◦ C.
(a) (b)
Figure 16. Experiment results for the battery pack characteristics: (a) discharge; and (b) charge.
Figure 17 shows the experimental results of the online SOC estimation using the ANN,
and the estimated SOC was compared with the reference SOC. As a result, the maximum
SOC errors in the discharge and charge processes were 2.3% and 2.6%, respectively. This
result has satisfied the initial design target of the SOC estimation error for the smart BMS.
145
Energies 2021, 14, 2634
(a) (b)
Figure 17. Experiment results of the SOC estimation using the ANN: (a) discharge; and (b) charge.
Next, the above experiment data including the battery voltage, current, and temper-
ature were used to calculate the SOC offline using the other two methods in the Matlab
simulation model. Figure 18 shows the comparison of the SOC estimation results using
three methods with the reference SOC during the discharge and charge processes of the
battery. The detailed absolute SOC errors of each method are given in Figure 19 and Table 6.
Using the EKF, the maximum SOC errors in the discharge and charge processes were
2.8% and 2.4%, respectively, which were similar to that of using the ANN. We found the
significant improvement for the SOC estimation by combining the EKF with ANN, which
had the SOC error of less than 1%. Comparisons were made with other SOC estimation
methods [37–49] and the maximum estimation errors are summarized in Table 7. Through
this, it was confirmed that the proposed method guarantees the accuracy of SOC estimation
similar to or better than other methods.
(a) (b)
Figure 18. Experiment results of the SOC estimations using three methods: (a) Discharge; and (b) Charge.
146
Energies 2021, 14, 2634
(a) (b)
Figure 19. Errors of the SOC estimations using three methods: (a) Discharge; and (b) charge.
Table 6. Summary of the errors in the SOC estimations using three methods.
Methods ANN Based EKF Based EKF+ANN Based

Maximum error 2.3% 2.8% 0.9%
Discharge
Average error 0.86% 0.77% 0.29%
Maximum error 2.6% 2.4% 0.8%
Charge
Average error 1.26% 1.53% 0.55%
Table 7. Comparison of the designed SOC estimation methods with others.
SOC Estimation Methods Maximum Estimation Errors

ANN based 2.6%
This study EKF based 2.8%
EKF+ANN based 0.9%
Fuzzy logic based 2.12% [37], 2.45% [38], 1.9% [39]
Impedance based 4% [40], 1% [41], 2.3% [42]
Other studies Unscented Kalman Filter based 1.22% [43], 1.5% [44]
Adaptive EKF based 2% [45], 2.5% [46], 1.6% [47]
Coulomb counting 3% [48], 4% [49]
4.3. Discussions
From the above experimental results, it can be seen that the SOC estimation methods
proposed in this study were accurate and satisfied the requirements of the designed smart
BMS. However, there are still issues to be investigated. The performances of the method
were evaluated when the battery pack was discharged and charged with a constant current.
In a real battery system such as an EV, the load current continuously changes according
to the vehicle speed. Therefore, to improve the quality of the battery modeling in the
EKF method, the dynamic characteristics of the battery must be considered. Applying
an ANN to EV requires more battery data trained with a dynamic current profile. In
addition, other important issues of SOC estimation and BMS design, such as cell balancing
and battery capacity fade, have not yet been considered. During long-term operation,
more experiments need to be performed to collect battery data for each cell and analyze
the cathode chemistry of the cells entirely. The next study will consider these issues and
improve the accuracy of the SOC estimation method for the LiBs in various real systems.
147
Energies 2021, 14, 2634
5. Conclusions
The authors proposed effective SOC estimation methods based on the EKF and ANN
for a LiB pack in a smart BMS. Detailed configurations and specifications of the smart
BMS and LiB pack were presented. First, an ANN was used to build an SOC estimation
model, which was trained and tested using a real battery dataset including voltage, current,
temperature, and measured SOC over 20 cycles. Inputs of the designed network consisted
of voltage, current, and temperature, and output was the estimated SOC of the battery. The
design process for the ANN was described in detail. The Google TensorFlow open-source
library was used to design and optimize the network configurations. Next, we developed
a SOC estimation algorithm using the EKF, in which the LiB model was studied and a
Thevenin model was developed to combine it with the Ah integration method. The current
and terminal voltage of the battery represent the input variables, and the SOC represents
the output variable. Finally, the EKF-ANN was proposed to improve the shortcomings of
the above two methods, where the ANN was redesigned by adding one more input of the
previous SOC determined using the EKF method. Both methods were confirmed through
experiments performed on real battery data collected from the battery pack consisting of
the LIB 18,650 35E cells at 4.2 V and 3.4 Ah. With the ANN and EKF, the SOC estimation
performances were almost similar with a maximum SOC errors of 2.4% to 2.8%. Meanwhile,
the use of the EKF-ANN significantly improved the accuracy of SOC estimation with less
than 1% error. We are confident that the results of this study can be effectively applied to a
smart BMS for industrial energy storage systems.
Author Contributions: Conceptualization, methodology: V.Q.D., M.-C.D. and C.S.K.; writing-

original draft preparation: V.Q.D.; experiment support: B.Z., C.-H.D., L.J. and M.-K.L.; validation,
writing-review and editing: M.-C.D. and B.Z.; project administration: J.H.B.; supervision: M.P. All
authors have read and agreed to the published version of the manuscript.
Funding: This research was supported by the “Development of high precision Smart-BMS for fire
prevention with HKF and Deep Learning” project of IES Co., LTD (Ho Chi Minh City, Vietnam) and
Korea Electrotechnology Research Institute.
Data Availability Statement: Data is contained within the article.
References
1. Khalid, M.; Sheikh, S.S.; Janjua, A.K.; Khalid, H.A. Performance validation of electric vehicle’s battery management system under
state of charge estimation for lithium-ion battery. In Proceedings of the 2018 International Conference on Computing, Electronic
and Electrical Engineering (ICE Cube), Quetta, Pakistan, 12–13 November 2018; pp. 1–5. [CrossRef]
2. Zheng, M.X.; Qi, B.J.; Du, X.W. Dynamic model for characteristics of Li-ion battery on electric vehicle. In Proceedings of the 2009
4th IEEE Conference on Industrial Electronics and Applications, Xian, China, 25–27 May 2009; pp. 2867–2871. [CrossRef]
3. Chen, T.M.; Jin, Y.; Lv, H.Y.; Yang, A.T.; Liu, M.Y.; Chen, B.; Xie, Y.; Chen, Y. Applications of Lithium-Ion Batteries in Grid-Scale Energy
Storage Systems; Transactions of Tianjin University: Tianjin, China, 2020; Volume 26, pp. 208–217. [CrossRef]
4. Mehr, T.H.; Masoum, M.A.S.; Jabalameli, N. Grid-connected Lithium-ion battery energy storage system for load leveling and
peak shaving. In Proceedings of the 2013 Australasian Universities Power Engineering Conference (AUPEC), Hobart, Australasia,
29 September 2013; pp. 1–6. [CrossRef]
5. Thorbergsson, E.; Knap, V.; Swierczynski, M.; Stroe, D.; Teodorescu, R. Primary Frequency Regulation with Li-Ion Battery Based
Energy Storage System-Evaluation and Comparison of Different Control Strategies. In Proceedings of the Intelec 2013 35th
International Telecommunications Energy Conference, Smart Power and Efficiency, Hamburg, Germany, 13–17 October 2013;
pp. 1–6.
6. Byrne, R.H.; Hamilton, S.; Borneo, D.R.; Olinsky-Paul, T.; Gyuk, I. The value proposition for energy storage at the sterling
municipal light department. In Proceedings of the 2017 IEEE Power & Energy Society General Meeting, Chicago, IL, USA,
16–20 July 2017; pp. 1–5. [CrossRef]
7. Hannan, M.A.; Hoque, M.M.; Hussain, A.; Yusof, Y.; Ker, P.J. State-of-the-Art and Energy Management System of Lithium-Ion Batteries
in Electric Vehicle Applications: Issues and Recommendations; IEEE: Piscataway, NJ, USA, 2018; Volume 6, pp. 19362–19378. [CrossRef]
148
Energies 2021, 14, 2634
8. Duan, J.; Tang, X.; Dai, H.; Yang, Y.; Wu, W.; Wei, X.; Huang, Y. Building Safe Lithium-Ion Batteries for Electric Vehicles: A Review.
Electrochem. Energ. Rev. 2020, 3, 1–42. [CrossRef]
9. How, D.N.T.; Hannan, M.A.; Hossain Lipu, M.S.; Ker, P.J. State of Charge Estimation for Lithium-Ion Batteries Using Model-Based and
Data-Driven Methods: A Review; IEEE: Piscataway, NJ, USA, 2019; Volume 7, pp. 136116–136136. [CrossRef]
10. Naguib, M.; Kollmeyer, P.; Emadi, A. Lithium-Ion Battery Pack Robust State of Charge Estimation, Cell Inconsistency, and Balancing:
Review; IEEE: Piscataway, NJ, USA, 2021. [CrossRef]
11. Mohamed, A.T. A review paper on batteries charging systems with the state of charge determination techniques. In Proceedings
of the 2nd Smart Cities Symposium (SCS 2019), Bahrain, Bahrain, 24–26 March 2019; pp. 1–6. [CrossRef]
12. Yuan, S.; Wu, H.; Yin, C. State of Charge Estimation Using the Extended Kalman Filter for Battery Management Systems Based on
the ARX Battery Model. Energies 2013, 6, 444–470. [CrossRef]
13. He, H.; Xiong, R.; Zhang, X.; Sun, F.; Fan, J. State-of-Charge Estimation of the Lithium-Ion Battery Using an Adaptive Extended
Kalman Filter Based on an Improved Thevenin Model. IEEE Trans. Veh. Technol. 2011, 60, 1461–1469. [CrossRef]
14. Chitnis, M.S.; Pandit, S.P.; Shaikh, M.N. Electric Vehicle Li-Ion Battery State of Charge Estimation Using Artificial Neural Network.
In Proceedings of the 2018 International Conference on Inventive Research in Computing Applications (ICIRCA), Coimbatore,
India, 11–12 July 2018; pp. 992–995. [CrossRef]
15. Xiong, R.; Cao, J.; Yu, Q.; He, H.; Sun, F. Critical Review on the Battery State of Charge Estimation Methods for Electric Vehicles; IEEE:
Piscataway, NJ, USA, 2018; Volume 6, pp. 1832–1843. [CrossRef]
16. Manthopoulos, A.; Wang, X. A Review and Comparison of Lithium-Ion Battery SOC Estimation Methods for Electric Vehicles. In
Proceedings of the IECON 2020 The 46th Annual Conference of the IEEE Industrial Electronics Society, Singapore, 18–21 October
2020; pp. 2385–2392. [CrossRef]
17. Chang, W.Y. The State of Charge Estimating Methods for Battery: A Review. Int. Sch. Res. Not. 2013, 2013, 953792. [CrossRef]
18. Zhang, R.; Xia, B.; Li, B.; Cao, L.; Lai, Y.; Zheng, W.; Wang, H.; Wang, W.; Wang, M. A Study on the Open Circuit Voltage and
State of Charge Characterization of High-Capacity Lithium-Ion Battery under Different Temperature. Energies 2018, 11, 2408.
[CrossRef]
19. Nugroho, A.; Rijanto, E.; Wijaya, F.D.; Nugroho, P. Battery state of charge estimation by using a combination of Coulomb
Counting and dynamic model with adjusted gain. In Proceedings of the 2015 International Conference on Sustainable Energy
Engineering and Application (ICSEEA), Bandung, Indonesia, 5–7 October 2015; pp. 54–58. [CrossRef]
20. Cuadras, A.; Kanoun, O. SoC Li-ion battery monitoring with impedance spectroscopy. In Proceedings of the 2009 6th International
Multi-Conference on Systems, Signals and Devices, Djerba, Tunisia, 23–26 March 2009; pp. 1–5. [CrossRef]
21. Spagnol, P.; Rossi, S.; Savaresi, S.M. Kalman Filter SoC estimation for Li-Ion batteries. In Proceedings of the2011 IEEE International
Conference on Control Applications (CCA), Denver, CO, USA, 28–30 September 2011; pp. 587–592. [CrossRef]
22. Rahmoun, A.; Biechl, H.; Rosin, A. SOC estimation for Li-Ion batteries based on equivalent circuit diagrams and the application
of a Kalman filter. In Proceedings of the 2012 Electric Power Quality and Supply Reliability, Tartu, Estonia, 11–13 June 2012;
pp. 1–4. [CrossRef]
23. Lee, S.J.; Kim, J.H.; Lee, J.M.; Cho, B.H. The State and Parameter Estimation of an Li-Ion Battery Using a New OCV-SOC Concept.
In Proceedings of the 2007 IEEE Power Electronics Specialists Conference, Orlando, FL, USA, 17–21 June 2007; pp. 2799–2803.
[CrossRef]
24. Yatsui, M.W.; Bai, H. Kalman filter based state-of-charge estimation for lithium-ion batteries in hybrid electric vehicles using
pulse charging. In Proceedings of the 2011 IEEE Vehicle Power and Propulsion Conference, Chicago, IL, USA, 6–9 September
2011; pp. 1–5. [CrossRef]
25. Haoran, L.; Liangdong, L.; Xiaoyin, Z.; Mingxuan, S. Lithium Battery SOC Estimation Based on Extended Kalman Filtering
Algorithm. In Proceedings of the 2018 IEEE 4th International Conference on Control Science and Systems Engineering (ICCSSE),
Wuhan, China, 21–23 August 2018; pp. 231–235. [CrossRef]
26. Yang, S.; Zhou, S.; Hua, Y.; Zhou, X.A.; Liu, X.H.; Pan, Y.W.; Ling, H.P.; Wu, B. A parameter adaptive method for state of charge
estimation of lithium-ion batteries with an improved extended Kalman filter. Sci. Rep. 2021, 11, 5805. [CrossRef] [PubMed]
27. Ismail, M.; Dlyma, R.; Elrakaybi, A.; Ahmed, R.; Habibi, S. Battery state of charge estimation using an Artificial Neural Network.
In Proceedings of the 2017 IEEE Transportation Electrification Conference and Expo (ITEC), Chicago, IL, USA, 22–24 June 2017;
28. Almeida, G.C.S.; Souza, A.C.Z.d.; Ribeiro, P.F. A Neural Network Application for a Lithium-Ion Battery Pack State-of-Charge
Estimator with Enhanced Accuracy. Proceedings 2020, 58, 33. [CrossRef]
29. Saji, D.; Babu, P.S.; Ilango, K. SoC Estimation of Lithium-Ion Battery Using Combined Coulomb Counting and Fuzzy Logic
Method. In Proceedings of the 2019 4th International Conference on Recent Trends on Electronics, Information, Communication
& Technology (RTEICT), Bangalore, India, 17–18 May 2019; pp. 948–952. [CrossRef]
30. Rivera-Barrera, J.P.; Muñoz-Galeano, N.; Sarmiento-Maldonado, H.O. SoC Estimation for Lithium-ion Batteries: Review and
Future Challenges. Electronics 2017, 6, 102. [CrossRef]
31. Liu, K.; Li, K.; Peng, Q.; Zhang, C. A brief review on key technologies in the battery management system of electric vehicles.
Front. Mech. Eng. 2019, 14, 47–64. [CrossRef]
149
Energies 2021, 14, 2634
32. Vaideeswaran, V.; Bhuvanesh, S.; Devasena, M. Battery Management Systems for Electric Vehicles using Lithium-Ion Batteries. In
Proceedings of the 2019 Innovations in Power and Advanced Computing Technologies (i-PACT), Vellore, India, 22–23 March
33. Ali, S.S.; Choi, B.J. State-of-the-Art Artificial Intelligence Techniques for Distributed Smart Grids: A Review. Electronics 2020,
9, 1030. [CrossRef]
34. Lai, X.; Qin, C.; Gao, W.; Zheng, Y.; Yi, W. A State of Charge Estimator Based Extended Kalman Filter Using an Electrochemistry-
Based Equivalent Circuit Model for Lithium-Ion Batteries. Appl. Sci. 2018, 8, 1592. [CrossRef]
35. He, H.; Xiong, R.; Fan, J. Evaluation of Lithium-Ion Battery Equivalent Circuit Models for State of Charge Estimation by an
Experimental Approach. Energies 2011, 4, 582–598. [CrossRef]
36. Chen, M.; Rincon-Mora, G.A. Accurate Electrical Battery Model Capable of Predicting Runtime and I–V Performance; IEEE Transactions
on Energy Conversion: Piscataway, NJ, USA, 2006; Volume 21, pp. 504–511. [CrossRef]
37. Jiani, D.; Zhitao, L.; Youyi, W.; Changyun, W. A fuzzy logic-based model for Li-ion battery with SOC and temperature effect. In
Proceedings of the 11th IEEE International Conference on Control & Automation (ICCA), Taichung, Taiwan, 18–20 June 2014;
38. Zheng, W.; Xia, B.; Wang, W.; Lai, Y.; Wang, M.; Wang, H. State of Charge Estimation for Power Lithium-Ion Battery Using a
Fuzzy Logic Sliding Mode Observer. Energies 2019, 12, 2491. [CrossRef]
39. Rao, R.P.; Bhat, R.S.; Ranjeeth, R.; Kavya, B.G. Implementing Fuzzy Logic to Improve the Accuracy of SoC Estimation for Li-ion
Battery. Int. J. Eng. Res. Technol. 2020, 09, 937–941.
40. Wu, S.; Chen, H.; Tsai, M.; Lin, T.; Chen, L. AC Impedance Based Online State-of-Charge Estimation for Li-ion Battery. In
Proceedings of the 2017 International Conference on Information, Communication and Engineering (ICICE), Xiamen, China,
17–20 November 2017; pp. 53–56. [CrossRef]
41. Xu, J.; Chris Mi, C.; Cao, B.G.; Cao, J.Y. A new method to estimate the state of charge of lithium-ion batteries based on the battery
impedance model. J. Power Sources 2013, 233, 277–284. [CrossRef]
42. Kim, T.; Qiao, W.; Qu, L. Real-time state of charge and electrical impedance estimation for lithium-ion batteries based on a hybrid
battery model. In Proceedings of the 2013 Twenty-Eighth Annual IEEE Applied Power Electronics Conference and Exposition
(APEC), Long Beach, CA, USA, 17–21 March 2013; pp. 563–568. [CrossRef]
43. Guo, X.; Xu, X.; Geng, J.; Hua, X.; Gao, Y.; Liu, Z. SOC Estimation with an Adaptive Unscented Kalman Filter Based on Model
Parameter Optimization. Appl. Sci. 2019, 9, 4177. [CrossRef]
44. Wang, W.; Wang, X.T.; Xiang, C.; Wei, C.; Zhao, Y. Unscented Kalman Filter-Based Battery SOC Estimation and Peak Power Prediction
Method for Power Distribution of Hybrid Electric Vehicles; IEEE: Piscataway, NJ, USA, 2018; Volume 6, pp. 35957–35965. [CrossRef]
45. Mazzi, Y.; Sassi, H.B.; Errahimi, F.; Es-Sbai, N. State of charge estimation using extended kalman filter. In Proceedings of the
2019 International Conference on Wireless Technologies, Embedded and Intelligent Systems (WITS), Fez, Morocco, 3–4 April
46. Ciortea, F.; Rusu, C.; Nemes, M.; Gatea, C. Extended Kalman Filter for state-of-charge estimation in electric vehicles battery packs.
In Proceedings of the 2017 International Conference on Optimization of Electrical and Electronic Equipment (OPTIM) & 2017 Intl
Aegean Conference on Electrical Machines and Power Electronics (ACEMP), Brasov, Romania, 25–27 May 2017; pp. 611–616.
[CrossRef]
47. Cheng, Z.; Lv, J.K.; Liu, Y.L.; Yan, Z.H. Estimation of State of Charge for Lithium-Ion Battery Based on Finite Difference Extended
Kalman Filter. J. Appl. Math. 2014, 2014, 10. [CrossRef]
48. Baccouche, I.; Mlayah, A.; Jemmali, S.; Manai, B.; Essoukri Ben Amara, N. $Implementation of a Coulomb counting algorithm for
SOC estimation of Li-Ion battery for multimedia applications. In Proceedings of the 2015 IEEE 12th International Multi-Conference
on Systems, Signals & Devices (SSD15), Mahdia, Tunisia, 16–19 March 2015; pp. 1–6. [CrossRef]
49. Qaisar, S.M. Event-Driven Coulomb Counting for Effective Online Approximation of Li-Ion Battery State of Charge. Energies
2020, 13, 5600. [CrossRef]
150
energies
Article
Simulation of Energy Exchange between Single Prosumer
Residential Building and Utility Grid
Andres Annuk 1, *, Wahiba Yaïci 2 , Matti Lehtonen 3 , Risto Ilves 4 , Toivo Kabanen 1 and Peep Miidla 5
1 Department of Energy Application Engineering, Institute of Technology, Estonian University of Life Sciences,
51006 Tartu, Estonia; toivo.kabanen@emu.ee
2 CanmetENERGY Research Centre, Natural Resources Canada, 1 Haanel Drive,
Ottawa, ON K1A 1M1, Canada; wahiba.yaici@canada.ca
3 Department of Electrical Engineering and Automation, Aalto University, 02150 Espoo, Finland;
matti.lehtonen@aalto.fi
4 Department of Biosystems Engineering, Institute of Technology, Estonian University of Life Sciences,
51006 Tartu, Estonia; risto.ilves@emu.ee
5 Estonian Center of Industrial Mathematics, 50090 Tartu, Estonia; peep.miidla@ut.ee
* Correspondence: andres.annuk@emu.ee; Tel.: +372-55-6820624
Abstract: Modern households usually have independent energy sources such as wind generators,
photovoltaic (PV) panels, and similar green energy production equipment. Experts predict that
soon, there will be an increasing number of such prosumers who both produce and consume energy.
This process alleviates and reduces the load on large national electricity networks and also contributes
to overall energy security. In this paper, a simulation model of a household, which employs a wind
generator as its independent source of electricity, is developed. It is expected that this approach will
be easily replicated for more complex configurations. The other components of the single prosumer
microgrid that will be assessed are the non-shiftable electricity consumption equipment, which is
Citation: Annuk, A.; Yaïci, W.; used mainly in households and deployed separately for water heater, with a separate battery to meet
Lehtonen, M.; Ilves, R.; Kabanen, T.; the needs of these non-shiftable consumers. The 5-min data intervals for the year of simulation have
Miidla, P. Simulation of Energy been used. The characteristics of energy flow according to production and consumption schedules
Exchange between Single Prosumer
and the capacity of storage equipment have been modelled and simulated. Results disclose that wind
Residential Building and Utility Grid.
turbine production size and buffer battery have a crucial impact on the demand cover factor.
Energies 2021, 14, 1553. https://
doi.org/10.3390/en14061553
Keywords: load shifting; energy storage; wind energy; green energy; self-consumption; cover factor;
Academic Editor: Andrei Blinov
microgrids; buffer battery; distributed generation; simulation
Received: 10 February 2021

Accepted: 8 March 2021
Published: 11 March 2021 1. Introduction
The European Union’s energy policy aims to achieve 32% of its total electricity pro-
Publisher’s Note: MDPI stays neutral duction from renewable sources by 2030 [1]. Increasingly, more attention is being paid to
with regard to jurisdictional claims in converting energy from renewable sources such as wind and solar [2]. These natural energy
published maps and institutional affil- sources present new challenges for electrical engineers and researchers regarding maximal
iations. extractions of energy for on-site use, and the methods of their deployment besides fossil en-
ergy sources [3]. The renewable sources are interfaced with the grid by the means of power
electronic converters. Such systems are becoming widespread and getting more efficient
with the developments in topologies and power semiconductor components [4,5]. In terms
Copyright: © 2021 by the authors. of national energy security, it is important to maintain enough rotating reserve because this
Licensee MDPI, Basel, Switzerland. helps to maintain the stability of the electricity grid [6,7]. Small cogeneration heat plants
This article is an open access article that are powered by biofuels can support grid stability to an extent, but the combustion of
distributed under the terms and biofuels on the other hand harms the environment [8]. Hydropower is environmentally
conditions of the Creative Commons friendly and stable in 24-h cycles, although it varies from season to season [9,10].
Attribution (CC BY) license (https:// Another important aspect of electricity use is storage. Storage is possible only at vary-
ing levels. Pumped hydropower plants can be used for large-scale electricity storage [11].
4.0/).

Energies 2021, 14, 1553
Flywheel storage [12] is also used as a storage device in smart grids, transportation and
for maintaining grid stability; however, for the most part, private use remains theoretical.
Europe largely supports the transition to minimize energy use in which most new build-
ings will consume nearly zero energy with autonomous energy production and adequate
storage systems. In the past decade, microgrid-based prosumers have grown exponentially,
and it is desired that these prosumers would play a more appreciable role in optimizing
the operations of utility grids [13–15].
This article provides a simulation model of a household with a wind generator as
its electricity source. The other components of microgrid under review are so-called non-
shiftable equipment for electricity consumption, which are used in every home. We consider
these appliances and household equipment, known as “non-shiftable” (NS) consumption,
in the energy scheme and simulation experiments. Water consumption and water tank
as storage are evaluated separately in the model because of the requirement for domestic
water to be preheated before use. The final component of the energy scheme within the
microgrid or household is a separate battery that ensures that the needs of NS consumers
are best met and which can store as much energy as possible when it is produced by the
wind generator.
Conventionally, energy consumed in households emanates from either the wind
generator [16] or the utility grid. To attain self-sufficiency, i.e., where electricity generated
locally matches local consumption, it is necessary to find an optimal combination of
installed electricity production units, and storage buffers to mitigate the volatility of the
primary energy carrier such as wind or solar irradiation [2]. The present study introduces
a buffer battery (BB), which stands between the microgrid and utility grid. BB belongs
to the household as material equipment but is virtually considered as energy, being one
of the stores for the energy produced by the wind generator. The energy from the wind
generator that is not consumed in the microgrid during the observation period is loaded
there. If necessary, the energy needed for consumption in the microgrid is also provided
from BB. The energy that cannot be contained in the BB is transmitted to the utility grid.
It is, thus, possible that the energy consumed in the household can come from either wind
generator, buffer or grid.
Cover factor augmentation is indirectly assessed by certain authors. An overview [17]
describes loss minimization and power quality in distributed grids and sets as objective
the decrease in active losses in batteries. Vanhoudt et al. [18] studied the possibility of
increasing self-consumption by heat pump, which is indirectly related to energy storage.
By comparing different renewable electricity sources, they found that the wind generator’s
yield is better compared to photovoltaic (PV) panels. Naval et al. [2] modelled the versatility
of electricity sources and related real-time electricity prices, with the wind as one of the
most suitable primary energy carriers. The combination of wind and solar generation in a
microgrid was studied in [17,19]. For a net-zero energy (NZE) hybrid microgrid, combined
wind/solar generation with intermediate storage was analyzed using the HOMER Pro
software (Homer Energy LLC, Boulder, USA) )suite [20]. A major disadvantage of this
software is the 1-h or longer averaging period that it requires. Therefore, a less granular,
self-developed MATLAB (The MathWorks, Inc., Kista, Sweden )model with an averaging
period of 5 min was applied in this current research.
The cover factor is an indicator of load-shifting technology and is meant to handle
volatile primary energy carriers such as wind and sunlight, by deploying intermediate
storage devices [21–26]. In addition to shifting, Eltanay et al. [23] prioritized loads by
dividing them into two major groups. In some sources, it is described as load matching
index (LMI) or load generation matching index (LGMI) [27,28]. In [29] the optimal storage
capacity for full ride-throughs was discussed. Households require power supplies from
the largest possible number of renewable sources to reduce payoff times [2,30]. Increasing
self-consumption not only levels load peaks [31] but also decreases costs on the electrical
energy import to an economically feasible point of 60% self-consumption level [32].
152
Energies 2021, 14, 1553
The numerous studies described above indicate the increasing attention to the subject
of the microgrid. However, despite this proliferation of studies trying to solve and predict
the energy exchange between prosumer residential building and utility grid, to the authors’
knowledge, there is still insufficient research into or investigation on optimal system design
of a microgrid system in terms of wind turbine and buffer battery sizes effects on the
system performance and demand cover factor. The main novelty of this work is using a
buffer battery between of microgrid and utility grid to increasing of self-consumption. It is
not yet met in the scientific literature.
Hence, the main goal of this work is to simulate and establish setup configurations
of BB between microgrid and utility grid to increase self-consumption of the prosumer.
It is important to note that for evaluation of effectiveness and distinguishing of these
configurations, a new cover factor was introduced and used. The microgrid parameters
were also set. The approach can easily be generalized and replicated in more complex
configurations. Modelling and numerical experiments were carried out in a MATLAB
environment.
With these objectives in mind, the remaining article is structured as follows: Section 2
describes the configuration of the modelled household. Section 3 gives an overview of the
initial data that are used in simulation experiments. Section 4 describes the simulation
setup, while Section 5 presents the main results. Finally, the main conclusions drawn in
this work are provided in Section 6.
2. System Setup
On the base of our computer simulation is a typical private household with an ad-
ditional buffer battery (BB) between the local microgrid and the utility grid. The term
microgrid refers to an electrical installation, which comprises local electricity generation,
loads, storage and utility grid connection by default [33].
In Figure 1, the system setup consists of two scenarios. In the first, the microgrid is
connected to the utility grid Figure 1a and the energy change between microgrid (MG) and
utility grid (UG) is direct and bidirectional. In Figure 1a, Arrow A denotes energy acquired
from UG, while Arrow B means excess energy produced by wind generator (WG) and
sold to UG. In the second scenario, the microgrid has external buffer battery (BB) storage
attached Figure 1b. In Figure 1b, Arrow C means the most expensive energy acquired from
UG for a house owner. Arrow D denotes energy, which is taken back to MG and this is free
for a homeowner as it was earlier saved to BB (Arrow E), produced by a wind generator,
and leftover from household consumption. Arrow F indicates the possibility that wind
generator works well and some energy goes to UG. Figure 2 presents MG. All processes
are driven by the load controller.
Utility grid Utility grid
Buffer battery
A B C D E
Microgrid Microgrid
ǻa) ǻb)
Figure 1. Energy flows between microgrid (MG) and utility grid (UG). (a) MG without BB, (b) MG
with (BB).
153
Energies 2021, 14, 1553
Utility Battery Wind

grid storage turbine
(a) (b), (c)
Load
controller
(d) (e)
Non-shiftable Water
load heater
Figure 2. Configuration of microgrid (MG). (a) bidirectional energy flow between load controller
and UG, (b) bidirectional energy flow between load controller and battery, (c) one directional energy
flow from WG to load controller, (d) one directional energy flow from load controller to NS, (e) one
directional energy flow from load controller to WH.
The microgrid described in this work can be connected to the energy hub. An energy
hub is considered a unit where multiple energy carriers can be converted, conditioned,
and stored [34]. It is to be noted that this microgrid output is very stochastic and has the
best cooperation with the UG. Through the UG there is also a possible connection with
other energy carriers.
3. Input Data
As input data, a time series of production of the WG which is scaled to rated power
Pnom = 5 kW, manufactured by TUGE Ltd (TUGE Ltd., Paldiski, Estonia) [35] was used.
It is located in a coastal area with coordinates N 59.087694, E 23.591719. The dataset
collected covers the period from 1 December 2015 to 30 November 2016, taking into
consideration that December is the first winter month, and facilitating further seasonal
analysis. The average WG power output is derived by dividing the electricity generated
during the last sampling period by the length of the sample period Δt. In the actual research,
Δt = 5 min, and a year is divided into intervals of 5 min in length. Raw consumption data
is measured at a frequency of 4 times per second with the network analyzer—Chauvin
Arnaux (Chauvin Arnaux Metrix, Paris, France). Unfortunately, raw data have not survived.
After the measurement, raw data have been averaged to 10 s of time series data. In the next
step, data were converted to 5 min averaged interval data. Production data was processed
from the 5 min averaging interval. Generation and loads are sampled equally.
Figure 3 shows the data for one week in December 2015 in 5 min periods of energy
units. Battery and WH capacities are considered usable net values. Adding supercapacitors
as levelling elements to decrease excess power from wind generator to batteries in our
case is not used because of the small probability to increase maximum power allowed
to batteries. Energy-related parameters such as WH temperature are not appraised, and
neither is the energy necessary to reach WH minimal temperature of 55 ◦ C, to avoid
proliferating Legionella bacteria [36]. Operating temperature is assumed to have been
attained, and the state-of-charge charge/discharge dependency is neglected.
154
Energies 2021, 14, 1553
Figure 3. A week sample of initial data from December 2015.
The annual output of the WG is scaled to match the annual load as closely as possible.
A typical private household’s load pattern is considered with non-shiftable (NS) and hot
water production (B) parts. The recorded annual electricity consumption was 3473 kWh,
47% (1632 kWh) being allocated to NS, and 53% (1841 kWh) to B, based on real measurement
data [37]. We aimed to test the developed model and algorithm from our measured data.
The Nordic climate conditions in Estonia are very changeable and even harsh given the
very cold weather conditions in winter. Climate conditions are also accompanied by
consumption data.
4. Simulation Model
In our simulation model, we first consider the control of energy flows in the power
supply system of a private house that is sourced from a wind generator and that also has
access to a utility grid. In Figure 2 we see the configuration of the microgrid (MG). The load
controller in the center regulates sharing of energy produced by the generator.
This model has been created based on the principle that has been introduced in the
sources [18,38]. This model belongs to multi-period multiple time scales over the year
type [39]. In selecting wind turbine production data, we have based our analysis on
one-year data with average wind conditions. Thus, the results of the simulation do not
reflect the results for the different years to be taken into account when using outputs.
This methodology does not command forecasting or economic issues [40]. These topics are
planned in the following studies.
The sharing algorithm in every time interval is as follows [41], the variables denote
(all in kWh):
X1 : is the output energy of wind generator in the current interval.
X2 : is the energy stored in a battery in the current interval.
Z1 : is the energy needed for NS consumption in the current interval.
Z2 : is the energy needed for hot water equipment in the current interval.
The simulation algorithm is delineated as follows:
(1) Next time interval is opened. Main action—the rest of the energies from previous
intervals is transferred to the current one.
(2) WG output X1 first satisfies NS loads Z1 .
155
Energies 2021, 14, 1553
(3) If X1 > Z1 , the quantity X1 –Z1 forwarded to WH. Steps 4 and 5 are skipped.
(4) If X1 < Z1 , the quantity Z1 –X1 is taken from the battery, if possible, i.e., if Z1 –X1 < X2 .
Otherwise, missing energy comes from UG.
(5) Energy Z2 is taken from a battery or UG.
(6) Energy movements saved.
If the process does not require all the energy produced in the current time interval,
the excess will move to the battery, and if there is still more energy left after that, it will go
to the utility grid.
For evaluation of effectiveness and distinguishing these configurations and energy
flows, this paper introduced and used a new cover factor. In literature, one can find
several forms of cover factors. In our numerical experiments, expressed in detail below,
the following formula was applied:
YD = (W1 + W2 + W3 + W4 )/Wtotal (1)
where:
1. YD is the demand cover factor.
2. W1 is the total annual amount of energy produced by WG, which is directly consumed
by NS devices needs.
3. W2 is the total annual amount of energy produced by WG, which is used with hot
water consumption from WH.
4. W3 is the total annual amount of energy produced by WG, which is used for NS
consumption from the battery during the year. If the WG is unable to supply directly
to NS load, the missing energy will be taken from the battery.
5. W4 is the total annual amount of energy flowing back from BB to MG. This is if there
is not enough power in the battery and WH, then the energy is taken from BB. This is
new in this paper. The flow is depicted as Arrow D in Figure 1b and is decisive for
finding the new demand cover factor value YD . If there is not enough energy in BB,
it is taken directly from UG Arrow C on Figure 1b.
6. Wtotal is the total annual amount of energy consumed for water heating and non-
shiftable load, in other words, the total energy consumption of the entire household.
The cover factor, shortly, is the ratio of energy produced by the wind generator,
which is consumed in the household under consideration. It is easy to see that the following
inequalities hold:
0 ≤ YD ≤ 1 (2)
Indeed, YD = 0 only when the wind generator exceptionally does not produce anything
during the year. Denote with W the total annual amount of energy produced by the wind
generator. For batteries in microgrid and BB, one can only load energy from a wind
generator. It means that energy flow through Arrow E on Figure 1b can be only from the
wind generator. The same holds for Arrow D and the same fact is true for MG battery.
YD has no unit.
We have, therefore, the following expression:
W1 + W2 + W3 + W4 ≤ W (3)
Because the energy amounts W1 , W2 , W3 , and W4 used for microgrid consumption

loads are produced by the same WG during the year. Mention that the initial conditions
for energy in storage devices are set to zero in simulations. As the consumption from WG
cannot exceed the total consumption inside MG then:
W1 + W2 + W3 + W4 ≤ Wtotal (4)
The last inequality holds because the total consumption load Wtotal may contain a
part of energy acquired from the utility grid as we see from Steps 4 and 5 of the algorithm.
156
Energies 2021, 14, 1553
According to the definition based on Equation (1) of cover factor YD , calculation from
Equation (4) concludes that YD ≤ 1. Therefore, the inequalities in Equation (2) are proved.
The microgrid solution we offer, together with the BB between MG and UG, is subject
to certain limitations. This solution is designed to meet the needs of a private house as it is
detailed in the above described methodology. This system is built on the principle that the
uncertain parameters of the WG output power are grounded in storage devices such as WH
and batteries. The present approach adopted in this work is mainly focused on component-
and system-level design approaches rather than taking system parameter uncertainty
modelling as done by the source [42] or by using robust optimal energy management [40].
There is no way to include electric car chargers in this system, nor, for example, fast boilers.
To do this, the microgrid must be built differently, given the larger instant consumer power.
5. Results and Discussions

To achieve the goals of this work, computer simulations were carried out. The whole
energy system of the household Figures 1 and 2 is inserted into a simulation model with
linear charge and discharge characteristics. The input variables for the model are the year-
long time series of wind generator output; non-shiftable loads; hot water consumption.
The length of all three time series is 105, 120 and this is also the number of time intervals
we used.
To estimate the coincidence between generation and load, the cover factor described
by Equation (1) is applied. Cover factor YD characterizes the local generation/demand
ratio. The simulation is based upon flow charts in Figure 1.
Based on Figure 4, the values of microgrid equipment parameters for numerical
simulation experiments have been fixed. Because of practical considerations and expert
assessment, the capacitance values of both the microgrid battery and WH are chosen to
be 6 kWh. In future work, we intend to find a numerical method for the quantitative
evaluation of these values and estimation of cover factor YD increments.
Figure 4. Cover factor YD dependency on battery and water heater (WH) capacity.
157
Energies 2021, 14, 1553
In the next stage of the simulation experiment, the amplification coefficient RS was
introduced to find the dependence in wind generator production of energy flow from
microgrid through BB to utility grid Arrow F in Figure 1b and vice versa Arrow A in
Figure 1a. The amplification coefficient (RS ) is the ratio of the energy produced per year
by the wind generator to the energy consumed in the microgrid. With different values
of coefficient RS , the output of the wind generator is multiplied. Figure 4 depicts energy
flows, where RS varies between 0.8 and 3. At the crossing point or collocation point of
two lines, RS = 1, (it means the original production time series), the absorbed and injected
energy become equal. Further increase of RS results in a linearly growing part of generated
electricity fed into the utility grid. The energy, absorbed from the UG, is characterized by a
slightly falling line.
In equilibrium Figure 5, the exported and imported energy equal both 1401 kWh,
with YD being = 0.597. Even a minor wind generator over-leverage results in significantly
more electricity being injected into the utility grid.
Figure 5. Dependence of energy flows between the utility grid (UG) and microgrid (MG) without buffer battery (BB) for
different RS values.
The next experiment examines the dependency of energy flows from the buffer battery
size. Figure 6 shows the direct energy flow from grid to MG Figure 1b, Arrow C. Figure 6
depicts the electricity flow back from BB to MG Figure 1b Arrow D. Considering BB
size 10 kWh in Figure 5, the grid-to-microgrid energy is equal to 1180 kWh, while BB-to-
microgrid energy becomes equal to 221 kWh Figure 6. These two numbers sum up as
1401 kWh, which is valid for the “bufferless” case Figure 5. The same approach can be
applied to other RS and BB values.
158
Energies 2021, 14, 1553
1600
1400
RS = 0.8
RS = 1.0
; RS = 1.2
<
1200 RS = 2.0
RS = 3.0
Energy, kWh
1000
800
600
400
200
0
0 10 20 30 40 50
Buffer size, kWh
Figure 6. Energy from the utility grid (UG) to microgrid—MG.
Figure 6 shows direct energy flow from UG to MG Figure 1b, (C). Figure 7 depicts the
electricity flowing back from BB to MG Figure 1b (D). Considering BB size 10 kWh Figure 5
the grid-to-microgrid energy is equal to 1180 kWh, while BB-to-microgrid energy becomes
equal to 221 kWh Figure 5. These two numbers sum up as 1.401 kWh, which is valid for
bufferless operation Figure 4. The same approach applies to other RS and BB values as
well. In Figure 7, energy amounts are in the y-axis on the Formula 2 fraction line the fourth
member W4 .
800
700
600
Energy, kWh
500
400
300 RS = 0.8
; RS = 1.0
< RS = 1.2
200 RS = 2.0
RS = 3.0
100
0
0 10 20 30 40 50
Buffer size, kWh
Figure 7. Energy from buffer battery (BB) to microgrid (MG).
159
Energies 2021, 14, 1553
Figure 8 depicts the energy flow from BB to UG. If RS = 1 when moving to UG energy
amount is 1164 kWh, which is nearly equal to MG entering energy flow, that is 1180 kWh,
as in Figure 5. It was found that this difference is not decisive. The small difference is caused
by RS = 1 overproduction of 100 kWh from consumption, which is due to scaling conditions.
8000
7000
RS = 0.8
6000 RS = 1.0
RS = 1.2
RS = 2.0
RS = 3.0
Energy, kWh
5000
4000
3000
2000
;
<
1000
0
0 10 20 30 40 50
Buffer size, kWh
Figure 8. Energy from buffer battery (BB) to the grid.
Increasing RS moves all residual energy to UG, as seen in Figure 8. In Figure 9, it can
be seen dependence from BB size to cover factor by different RS . The reasonable capacity
of BB is 10 kWh. If BB = 0 when by RS = 1 is YD = 0.597 and by BB = 10 kWh YD = 0.66.
If we have RS = 2, when YD = 0.796. As seen in Figure 9, the cover factor increase occurs
when RS is much bigger than the BB size, but BB size multiplies RS influence.
Figure 9. Influence of BB size and RS on cover factor.
Based on Figure 9, results reveal that both, the BB and RS have an obvious influence
on the demand cover factor. This suggests that the greater capacity factor of WG causes an
additional increment in YD .
6. Conclusions
This paper developed and presented a simulation model of energy flows to investigate
the self-consumption of a household with a wind generator as its independent source of
160
Energies 2021, 14, 1553
electricity. The main assessment parameter is the demand cover factor (1). As input data,
time series of typical private household load patterns with non-shiftable (NS) and hot
water consumption (B) parts Figure 3, as well as the time series of production of the wind
generator (WG) with rated power Pnom = 5 kW, were used. This simulation approach is
easily generalized/replicated in more complex configurations of households, and also for
different individual places such as warehouses, industrial buildings, etc. By increasing
local consumption from renewable energy sources, energy losses in the utility grid and
greenhouse gas emissions are reduced.
The study demonstrated that a buffer battery introduced between the utility grid and
household inner microgrid has a notably positive influence on the demand cover factor.
As a result, it is possible to reduce the amount of energy purchased from the utility grid,
as seen in Figures 6 and 7. Figure 6 shows that if BB = 10 kWh, RS = 1, is purchased
from UG to MG 1180 kWh. If RS = 2 and BB= 10 kWh, we only need to buy 701 kWh of
electricity from UG. Wind turbine oversizing is reasonable based on these calculations,
as wind production varies greatly over the years. Figure 7 shows the amounts showing
how much less electricity we buy from UG compared to BB = 0.
It was proved that the production of the wind turbine is crucial for the value of the
demand cover factor. To demonstrate this, the amplification coefficient RS was introduced
to find the dependency of wind generator production of energy flow from microgrid to the
utility grid and vice versa. Figure 9 shows that compared to the baseline situation as RS = 1
and BB = 0 YD = 0.597 then in a situation where RS = 2 and BB = 10 kWh are YD = 0.796.
This is a significant increase.
Moreover, the amplification coefficient RS affects the cover factor only when applied
to wind generator production. Numerical experiments showed that when applied to
consumption data, the change is marginal.
Future research may focus on the WG production forecast and economic factors and
parameters, which should be included in real simulation models.
Author Contributions: A.A. supervised the work and prepared the original research; W.Y. and M.L.
provided the support in formulating the simulation problem; R.I. and T.K. draft preparation and
evaluation of manuscript; P.M. conceptualization and MATLAB simulation. All authors have read
and agreed to the published version of the manuscript.
Funding: This research was funded by the Estonian Centre of Excellence in Zero Energy and Resource
Efficient Smart Buildings supported this research and Districts, ZEBE, grant TK146 funded by the
European Regional Development Fund.
Acknowledgments: The authors would like to thank Tuge Energia Ltd., for making the data of the
10 kW wind generators available, Alo Allik for consumption data for this study and the Estonian
Centre of Excellence in Zero Energy and Resource Efficient Smart Buildings and Districts, ZEBE,
grant TK146 funded by the European Regional Development Fund supported this research.
References
1. Energy Efficiency Directive. Available online: https://ec.europa.eu/energy/topics/energy-efficiency/targets-directive-and-
rules/energy-efficiency-directive_enstrategy (accessed on 10 January 2021).
2. Naval, N.; Sánchez, R.; Yusta, J.M. A virtual power plant optimal dispatch model with large and small-scale distributed renewable
generation. Renew. Energy 2020, 151, 57–69. [CrossRef]
3. Annuk, A.; Jõgi, E.; Hovi, M.; Märss, M.; Uiga, J.; Hõimoja, H.; Peets, T.; Kalder, J.; Algirdas, J.; Allik, A. Increasing self electricity
consumption by using double water heating tanks for residential net zero energy buildings. In Proceedings of the 6th International
Conference on Renewable Energy Research and Application (ICRERA), San Diego, CA, USA, 5–8 December 2017; Volume 6,
4. Chub, A.; Zdanowski, M.; Blinov, A.; Rabkowski, J. Evaluation of GaN HEMTs for high-voltage stage of isolated DC-DC
converters. In Proceedings of the 10th International Conference on Compatibility, Power Electronics and Power Engineering
(CPE-POWERENG), Bydgoszcz, Poland, 29 June–1 July 2016; pp. 375–379. [CrossRef]
161
Energies 2021, 14, 1553
5. Blinov, A.; Chub, A.; Vinnikov, D.; Rang, T. Feasibility study of Si and SiC MOSFETs in high-gain DC/DC converter for renewable
energy applications. In Proceedings of the IECON 2013—39th Annual Conference of the IEEE Industrial Electronics Society,
Vienna, Austria, 10–14 November 2013; pp. 5975–5978. [CrossRef]
6. Lepa, J.; Annuk, A.; Kokin, E.; Põder, V.; Jürjenson, K. Energy production and consumption charts in energy system. Oil Shale
2009, 26, 309–318. [CrossRef]
7. Zemug, L.; Su, S.; Yunning, Z.; Xiaolomg, J.; Houhe, C.; Yujiang, R. Energy management strategy of active distribution network
with integrated distributed wind power and smart buildings. IET Renew. Power Gener. 2020, 14, 2255–2267. [CrossRef]
8. DeCicco, M.; Liu, D.Y.; Heo, J.; Krishnan, R.; Kurthen, A.; Wang, L. Carbon balance effects of U.S. biofuel production and use.
Clim. Change 2016, 138, 667–680. [CrossRef]
9. Okok, M.O.; Ruchathi, G.; Oromat, E. Expanding access to clean energy in developing countries: The role of off-grid mini hydro
power projects in Kenya. Int. J. Renew. Energy Res. 2019, 9, 1571–1577.
10. Hadiyanto, D.; Purwanto, Y.A.; Noorachmat, B.P.; Sapei, A. An indicator and evaluation criteria for off-grid micro-hydro power
sustainability assessment. Int. J. Renew. Energy Res. 2019, 9, 1–15.
11. Zoss, T.; Karklina, I.; Blumberga, D. Power to gas and pumped hydro storage potential in Latvia. Energy Procedia 2016, 95, 528–535.
[CrossRef]
12. Kesgin, M.G.; Han, P.; Taran, N.; Ionel, D.M. Overview of flywheel systems for renewable energy storage with a design study for
high-speed axial-flux permanent-magnet machines. In Proceedings of the 8th International Conference of Renewable Energy
Research and Applications (ICRERA) 2019, Brasov, Romania, 3–6 November 2019; pp. 1026–1031. [CrossRef]
13. Rehman, Z.; Mahmood, A.; Razaq, S.; Wamiq, A.; Naeem, U.; Shehzad, K. Prosumer based energy management and sharing in
smart grid. Renew. Sustain. Energy Rev. 2018, 82, 1675–1684. [CrossRef]
14. Jie, C.; Bu, Z.; Wang, Y.; Yang, H.; Jiang, J.; Li, H.-J. Detecting prosumer-community groups in smart grids from the multiagent
perspective. IEEE Trans. Syst. Man Cybern. Syst. PP 2019, 99, 1–13. [CrossRef]
15. Bautista-Villalon, M.; Gutirerrez-Villalobos, J.; Rivas-Araiza, E.A. IoT-based system to monitor and control household lighting
and appliance power consumption and water demand. In Proceedings of the 7th International Conference of Renewable Energy
Research and Applications (ICRERA) 2018, Paris, France, 14–17 October 2018; pp. 785–790.
16. Blinov, A.; Vinnikov, D.; Husev, O.; Chub, A. Experimental analysis of wide input voltage range qZS-derived push-pull
DC/DC converter for PMSG-based wind turbines. In Proceedings of the PCIM Europe, Nuremberg, Germany, 14–16 May 2013;
pp. 1435–1444.
17. Sambaiah, K.S. A review on optimal allocation and sizing techniques for DG in distribution systems. Int. J. Renew. Energy Res.
2018, 8, 1236–1256.
18. Vanhoudt, D.; Geysen, B.; Claessens, F.; Leemans, L.; Jespers, L.; Bael, J.B. An actively controlled residential heat pump: Potential
on peak shaving and maximization of self-consumption of renewable energy. Renew. Energy 2014, 63, 531–543. [CrossRef]
19. Annuk, A.; Jõgi, E.; Lill, H.; Kalder, J.; Hovi, M.; Pihlap, H.; Härm, M.; Allik, A. Augmentation of self-consumption of electricity
by using boilers and batteries for residential buildings. In Proceedings of the 7th International Conference of Renewable Energy
Research and Applications (ICRERA), Paris, France, 14–17 October 2018; Volume 7, pp. 256–260. [CrossRef]
20. Budes, E.E.B.; Ochoa, G.V.; Obregon, L.G.; Arango-Manrique, A.A.; Alvarez, J.R.N. Energy, economic, and environmental
evaluation of a proposed solar-wind power on-grid system using HOMER Pro® : A case study in Colombia. Energies 2020,
21. Katsaprakakis, D.A.; Zidianakis, G.; Yiannakoudakis, Y.; Manioudakis, E.; Dakanali, I.; Kanouras, S. Working on buildings’ energy
performance upgrade in mediterranean climate. Energies 2020, 13, 2159. [CrossRef]
22. Hossain, E.; Zawad, M.; Rakibul, K.H.; Islam, K.H.R.; Akash, M.Q. Design a novel controller for stability analysis of microgrid by
managing controllable load using load shaving and load shifting techniques; and optimizing cost analysis for energy storage
system. Int. J. Renew. Energy Res. 2016, 6, 772–786.
23. Eltamaly, A.M.; Mohamed, M.A.; Alolah, A.I. A novel smart grid theory for optimal sizing of hybrid renewable energy systems.
Sol. Energy 2016, 124, 26–38. [CrossRef]
24. Hassanzadeh, M.N.; Fotuhi-Firuzabad, M.; Safdarian, A. Wind energy penetration with load shifting from the system well-being
viewpoint. Int. J. Renew. Energy Res. 2017, 7, 977–987.
25. Duerr, S.; Ababei, C.; Ionel, D.M. Load balancing with energy storage systems based on co-simulation of multiple smart buildings
and distribution networks. In Proceedings of the 8th International Conference on Renewable Energy Research and Applications
(ICRERA) 2017, San Diego, CA, USA, 5–8 December 2017; Volume 5, pp. 175–180.
26. Jonaitis, A.; Miliute, R.; Deveikis, T. Dynamic model of wind power balancing in hybrid power system. Turk. J. Electr. Eng.
Comput. Sci. 2017, 25, 222–234. [CrossRef]
27. Bashir, A.A.; Kasmaei, M.P.; Safdarian, A.; Lehtonen, M. Matching of local load with on-site PV production in a grid-connected
residential building. Energies 2018, 11, 2409. [CrossRef]
28. Degefa, M.Z.; Lehtonen, M.; McCulloch, M.; Nixon, K. Real-time matching of local generation and demand: The use of high
resolution load modeling. In Proceedings of the IEEE PES Innovative Smart Grid Technologies Conference Europe, Ljubljana,
Slovenia, 9–12 October 2016.
29. Hauer, I.; Balischewski, S.; Ziegler, C. Design and operation strategy for multi-use application of battery energy storage in wind
farms. J. Energy Storage 2020, 32, 101572. [CrossRef]
162
Energies 2021, 14, 1553
30. Li, X.; Lim, M.K.; Ni, D.; Zhong, B.; Xiao, Z.; Hao, H. Sustainability or continuous damage: A behaviour study of prosumers´
electricity cinsumption after installing household disreibuted energy resources. J. Clean. Prod. 2020, 264, 121471. [CrossRef]
31. Rogeau, A.; Barbier, T.; Girard, R.; Kong, N. Evoluotion of electrical distribution grid sizing considering self-consunption of
local renewable producton. In Proceedings of the 24th International Confeence & Exhibition on Electricity Distribution (CIRED),
Glasgow, UK, 12–15 June 2017; pp. 1–5.
32. Dancker, J.; Götze, J.; Schultz, E.; Könneke, N.; Beyrau, F.; Wolter, M. Optimal design and operation of a CHP based district
heating system including a heat storage and electrode boiler to increase self-consumption. In Proceedings of the 2019 IEEE
Inovative Smart Grid Technologies Conference—Latin America (ISGT Latin America), Gramado, Brazil, 15–18 September 2019.
[CrossRef]
33. Kroposki, B.; Basso, T.; DeBlasio, R. Microgrid standards and technologies. In Proceedings of the IEEE Power and Energy Society
2008 General Meeting: Conversion and Delivery of Electrical Energy in the 21st Century, PES, Pittsburg, PA, USA, 20–24 July 2008.
[CrossRef]
34. Geidl, M.; Koeppel, G.; Favre-Perrod, P.; Klöckl, B.; Andresson, G.; Fröhlich, K. Energy hubs for the future. IEEE Power Energy Mag.
2007, 5, 24–30. [CrossRef]
35. TUGE 10 Technical Data, 10 kW. Tuge Energia Ltd., Paldiski, Estonia, 2020; pp. 4–6. Available online: http://www.tuge.ee/
products/tuge10 (accessed on 10 January 2021).
36. Volker, S.; Kistemann, T. Field testing hot water temperature reduction as an energy-saving measure—Does the Legionella
presence change in a clinic’s plumbing system? Environ. Technol. 2015, 36, 2138–2147. [CrossRef]
37. Annuk, A.; Hovi, M.; Kalder, J.; Märss, M. Consumption and Wind Production Time Series Data. Estonian University of Life
Sciences: Tartu, Estonia, 2020. Available online: https://doi.org/10.15159/eds.dt.20.02 (accessed on 10 January 2021).
38. Baetens, R.; De Coninck, R.; Roy, J.V.; Verbuggen, B.; Driese, J.; Helsen, L.; Saelens, D. Assessing electrical bottlenecks at feeder
level for residential net zero-energy buildings by integrated systemsimulation. Appl. Energy 2012, 96, 74–83. [CrossRef]
39. Sperstad, I.B.; Korpås, M. Energy storage scheduling in distribution systems considering wind and photovoltaic generation
uncertainties. Energies 2019, 12, 1231. [CrossRef]
40. Hossein, S.M.; Carli, R. A residential demand-side management strategy under nonlinear pricing based on robust model
predictive control. In Proceedings of the 2019 IEEE International Conference on Systems, Man and Cybernetics (SMC), Bari, Italy,
6–9 October 2019; pp. 3243–3248. [CrossRef]
41. Annuk, A.; Hovi, M.; Kalder, J.; Kabanen, T.; Ilves, R.; Märss, M.; Martinkauppi, B.; Miidla, P. Methods for increasing shares of
self-consumption in small PV solar energy applications. In Proceedings of the 9th International Conference on Renewable Energy
Research and Applications (ICRERA) 2020, Glasgow, UK, 26–29 September 2020; Volume 5, pp. 184–188. [CrossRef]
42. Shi, R.; Li, S.; Zhang, P.; Lee, K. Integration of renewable energy sources and electric vehicles in V2G network with adjustable
robust optimization. Renew. Energy 2020, 153, 1067–1080. [CrossRef]
163
MDPI
St. Alban-Anlage 66
4052 Basel
Switzerland
Tel. +41 61 683 77 34
Fax +41 61 302 89 18
www.mdpi.com
Energies Editorial Office

E-mail: energies@mdpi.com
www.mdpi.com/journal/energies
MDPI
St. Alban-Anlage 66
4052 Basel
Switzerland
Tel: +41 61 683 77 34
www.mdpi.com ISBN 978-3-0365-5278-1

Energy Management

Uploaded by

Copyright:

Available Formats

Energy Management

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Energy Management

Uploaded by

Copyright:

Available Formats

Power Electronics and

ISBN 978-3-0365-5277-4 (Hbk)

About the Editors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii

Omar Abdel-Rahim, Andrii Chub, Andrei Blinov, Dmitri Vinnikov and

Sumukh Surya, Akash Samanta, Vinicius Marcis and Sheldon Williamson

Zbigniew Rymarski, Krzysztof Bernacki and Łukasz Dyga

Akash Samanta and Sheldon S. Williamson

1 Department of Electrical Power Engineering and Mechatronics, Tallinn University of Technology,

Energies 2022, 15, 4550. https://doi.org/10.3390/en15134550 1 https://www.mdpi.com/journal/energies

2. Proposed Buck-Boost Dc–Dc Converter

Figure 1. Proposed buck-boost dc–dc converter.

2.1. Continuous Conduction Mode

where v L1 , v L2 , Vdc , Vo , IL , iC , id , M, and D are inductor L1 voltage, inductor L2 voltage,

2.2. Discontinuous Conduction Mode

Figure 4. Typical converter waveforms in DCM.

The duty cycle D2 is an unknown, so a second equation is needed to eliminate D2 .

ISN VGF/L· DT6 ISN VGF/L· DT6

The peak diode current could be obtained from the graph as

Equation (22) could be rearranged to

2.3. Switches’ and Diodes’ Voltage Stresses

VdsS1 = Vdc (25)

Figure 6. CCM and DC boundary.

The current stress of switch S1 is given by

Switches S2 and S3 face similar voltage and current stress, as follows:

Vd1 = Vdc (29)

2.4. Components’ Design

2.4.1. Inductors’ Design

&KDUJH 'LVFKDUJH &KDUJH

2.4.2. Capacitors’ Design

of f (Vo − IL)(1 − D ) TS Vo DTS

2.4.3. Comparative Study

ͳͷͲǡͲǤͷ ε ͳͷͲǡͳ ε ͳͷͲǡͷ ε ͳͷͲǡ͵

ͲǤʹ ͲǤ͵ ͲǤͶ ͲǤͷ ͲǤ͸ ͲǤ͹ ͲǤͺ

    

      

Table 1. Voltage gain and component stress comparison.

9N: ε 9N: ε 9N: ε 9N:

   

       

Table 2. Hardware prototype speciﬁcations.

1RQ,QYHUWLQJ 7UDGLWLRQDO 3URSRVHG

      

7UDGLWLRQDO 1RQ,QYHUWLQJ 3URSRVHG

 9LQ  9 6WHSXSUDWLR

&RQYHQWLRQDO 3URSRVHG 1RQ,QYHUWLQJ

       

1 Robert Bosch Engineering and Business Solutions, Bangalore 560100, India

Publisher’s Note: MDPI stays neutral 1. Introduction

Energies 2022, 15, 623. https://doi.org/10.3390/en15020623 19 https://www.mdpi.com/journal/energies

the expense of a high computational cost. Resistor–capacitor-based models can be opti-

2. Thermal Modeling and Temperature Estimation Strategy

2.1. Heat Generation Model

Figure 1. 1-RC ECM (Thevenin’s equivalent) model of an LIB cell.

2.2. First-Order Thermal Modelling

Figure 2. First-order heat resistor–capacitor-based thermal model of LIB.

2.2.2. KF for First-Order Thermal Model

2.3. Second-Order Thermal Modelling

Figure 3. Second-order equivalent circuit thermal model of LIB.

2.3.2. KF for Second-Order Thermal Model

Xk = Ak−1 Xk−1 + Bk−1 Uk−1 + Wk−1 (16)

ISN VGF/L· DT6 ISN VGF/L· DT6

ͳͷͲǡͲǤͷ ε ͳͷͲǡͳ ε ͳͷͲǡͷ ε ͳͷͲǡ͵

9N: ε 9N: ε 9N: ε 9N:

1RQ,QYHUWLQJ 7UDGLWLRQDO 3URSRVHG

7UDGLWLRQDO 1RQ,QYHUWLQJ 3URSRVHG

9LQ 9 6WHSXSUDWLR

&RQYHQWLRQDO 3URSRVHG 1RQ,QYHUWLQJ