INTERNATIONAL JOURNAL of RENEWABLE ENERGY RESEARCH

A. Patidar et al., Vol.14, No.2, June, 2024

Design and Development of Solar Based Fast

Charger for E-Vehicle Using MPPT Algorithm

Aayush Patidar* , Vijay Kumar Dubey**‡ , Dipali Sarvate***

*Department of Electrical Engineering, Prestige Institute of Engineering Management and research, Indore, India
** Department of Electrical Engineering, Prestige Institute of Engineering Management and research, Indore, India
*** Department of Electrical Engineering, Prestige Institute of Engineering Management and research, Indore, India
(ayupatidar@gmail.com, vijaydubeyy@gmail.com, dsarvate@piemr.edu.in )

‡
Corresponding Author; Aayush Patidar, Prestige Institute of Engineering Management and research, Indore, India,
Tel: +91 7509781787, ayupatidar@gmail.com

Received: 17.07.2023 Accepted:12.08.2023

Abstract- This paper relates with the Solar based fast charger for e-Vehicle. Currently, Solar Energy based e-vehicle charging
make drastic change in dependence on fossil fuel. Charging of e-vehicle are ideally assumed within 2-4 minutes like gas
station but in reality, it took more than a hour to fully charge a battery. The paper give an design solution for MPPT based fast
charger, focus mainly on the charging rate, energy efficiency and the reduction in power consumption compare to the
traditional street e- vehicle charging system. This paper part covered in MPPT Algorithms and Atmega328P microcontroller
having the function of Complementary Wave Generator which has been utilized for producing the High resolution PWM signal
that further operating the Synchronous Buck Converter. The converter circuit contain an extensive novel interlocking method
for protection against Overvoltage, Under Voltage Protection, and Reverse Current & Opposite Polarity. Energy efficiency
plays a crucial role in this paper, with particular emphasis on the dusk to dawn voltage as a significant parameter. To address
this, several modifications have been made from a hardware design perspective. However, incorporating these protective
measures poses challenges when implementing the maximum power point tracking algorithm to extract the maximum power
from the solar panel. The MATLAB 2016A software has been utilized to develop the proposed model, and the simulation
results and related circumstances are detailed in the result and discussion section.
Keywords DC-DC converter; Maximum Power Point Tracking; Photovoltaic System; Advance converter topology; Start up
and Running interlocks.
energy sources. However, an essential aspect is efficient
1. Introduction
power conversion from the energy source to storage devices
or loads [1]. Stand-alone applications have emerged as
All nations share a fundamental requirement to access
highly profitable systems from an energy supply perspective.
Sustainable and Cost-effective energy, which is crucial for
Several PV module integrated converters have been
their economic and Manufacturing expansion, as well as
developed for high efficiency in stand-alone systems [3].
societal and human growth. India holds the third position Micro-grid applications are also successfully meeting energy
globally as both a major producer and consumer of energy. requirements in various villages across India. Notably, when
According to records from the National Electric Grid, India implementing the CCM (Continuous Conduction Mode)
had already installed and operationalized approximately 370 operation in a converter that transfers energy from source to
GW of capacity by March 31, 2020. In the fiscal year 2018- load, stand-alone applications play a critical role in
19, India generated an estimated 1372 TWh of electrical determining the switching frequency boundary condition
energy for utility areas and 1547 TWh when non-utilities [3][5].
were included. The per capita gross electrical energy
generation in 2018-19 stood at 1191 kWh. To fulfill this purpose, various methods are employed for
solar energy harvesting. For instance, the Fractional Open
India has various sources of electrical supply, as depicted in Circuit Voltage method [2] considers the ratio of the
Fig.1, presenting ample opportunities in non-conventional
INTERNATIONAL JOURNAL of RENEWABLE ENERGY RESEARCH
A. Patidar et al., Vol.14, No.2, June, 2024

maximum peak voltage to the open-circuited voltage to

control the system. Additionally, other algorithms have been
developed and practiced in recent years, such as the Perturb
and Observe (P&O) method [4][5]. Although this method
operates efficiently, it exhibits significant output oscillations
at the maximum power point (MPP). To address this issue,
the Incremental Conductance algorithm [5][10] has been
employed, which utilizes variations in the conductance of the
PV cell. The Prediction Model-Guided Jaya Algorithm [11],
utilized extensively over the past 5 to 6 years, assists in
determining the exact MPP. The Fuzzy Logic algorithm [8]
is also a popular method, particularly when working with a
digital controller, as it effectively regulates the converter
under partially loaded conditions [9].
The market offers a range of lithium-ion cell variants,
including small cylindrical cells (with a solid body and no Fig. 1. Simplified Architecture of Hardware
terminals), large cylindrical cells (with a solid body and
terminals), flat or pouch cells (featuring a soft, flat body), The figure also presents the general model of a solar cell,
and rigid plastic case cells with large threaded terminals. The which is based on a single diode model. The expression for
latter two variants are sometimes referred to as prismatic this single diode model is as
cells due to their rectangular shape. These batteries exhibit
specific energy densities ranging from 100 to 250 Wh/kg,
volumetric energy densities ranging from 250 to 680 Wh/L,
and specific power densities ranging from 300 to 1500 W/kg. RSP – The Resistance in Parallel
Lithium-ion cells incorporating a positive electrode of
lithium iron phosphate and a negative electrode of graphite RSS – The Resistance in Series
typically exhibit a nominal open circuit voltage of 3.7 volts IPV –Photocurrent
and a typical charging voltage of 3.6 volts. They possess a
volumetric energy density of 220 Wh/L (790 kJ/L) and a VD – The Diode Voltage
gravimetric energy density of over 90 Wh/kg. These batteries
a – Ideality Factor
have a cycle life of 2000-7000 cycles to 80% of the original
capacity at 100% depth of discharge (DOD), and over 10,000 VT – Thermal Voltage
cycles to 80% of the original capacity at 10% DOD. The
But if many cells are taken in series and parallel, then the
cathode composition consists of approximately 90% C
equivalent expression for determining the photocurrent is as
LiFePO4 (phos-dev-12 grade), 5% carbon (superior
follow
graphite), and 5% poly-vinylidene fluoride. The cell
configuration comprises a carbon-coated aluminium current.

2. System Design

The system comprises a 1000W Solar Panel, a LiFePO4

Where NP is the no. of cells in Parallel & NS is the no. of
battery with a nominal voltage of 12.8V, and a load
cells in Series.
consisting of a e-vehicle charger. The simplified hardware
architecture is illustrated in Fig.1. 2.2 ATmega328P Microcontroller:
2.1 Solar Panel: This microcontroller is an 8 – bit, 28 – Pin Flash High
Performance RISC CPU based 32KB ISP Flash memory
For experimental purposes, a 60W Mono-crystalline Solar
with read-while-write capability. It has C-compiler optimized
Panel from CENTYSIS is incorporated into the system. The
architecture, 20MHz Internal Oscillator. We get the feature
solar panel specifications include an Open Circuit Voltage of
of Software Selectable Frequency range from 20MHz to 31
22.2V, a Short Circuit Current of 3.56A, an efficiency of
KHz, 10 – Bit 6 channel ADC pins,1K bytes EEPROM, 2 K
14.22%, a current temperature coefficient of 0.06, and a
bytes SRAM,4 PWM Generating PINs, 1 Complementary
voltage temperature coefficient of -0.34. Fig.4 displays the I-
Wave Generator Pin, & with ab programmable watchdog
V (current-voltage) and P-V (power-voltage) characteristics
timer with internal oscillator. Most of the PIN configured in
of the solar panel. At an irradiance level of 1000 W/m2, the
this project, its Wide Operation Voltage Variant ‘F’
PV cell can generate an output of over 1000W.
recommended from 2.3 V to 5.5 V whereas extreme Low
Power (XLP) ‘LF’ Variant is 1.8V – 5.5V. Basically we have
objective to implement whole operation in single system.
The controller has to work for measuring the voltage and
current which will provide input to MPPT algorithm [17] and

364
INTERNATIONAL JOURNAL of RENEWABLE ENERGY RESEARCH
A. Patidar et al., Vol.14, No.2, June, 2024

then it has to operate in function of Bulk Charge Mode,

Absorption Mode & then

Fig. 2. Buck Converter from PV cell to 12.8 LPF Battery

float mode. To ensure the long-term reliability of the system,

in addition to leveraging the operational features mentioned,
a safety circuit is implemented. This circuit is responsible for
maintaining the system's reliability. All input/output (IO)
ports of the system have a source/sink rating of 25 mA.
There are three timers in the system: TMR0 and TMR2 are The resistance R44 and R43 are feeding the required power
8-bit timers, while TMR1 is a 16-bit timer. This signal to the M1 and M2 MOSFET due to which the circuit
microcontroller is having special feature which we discussed get into the operation mode and battery start charging but
earlier that is Internal Oscillator that is the reason we don’t MOSFET U32 makes VGS= 0 due to which circuit isolate
need to configure the microcontroller through external for some time. This feature of circuit used with MPPT
oscillator circuit. This increases the overall reliability of the algorithm to achieve the goal. The MOSFET U28 is the
system. device which is getting the PWM signal from the
microcontroller while there’s another diode D31 is 1N4007
2.3 LM2950 Voltage Regulator IC:
ZENER diode which is working in the system to provide the
The LM2950 provided constant voltage to Microcontroller over-charge protection.
and other part of control circuit initially during start up of
2.5 Driver Circuit:
System. This IC provides 1V to 5V power signal to
controller. The other two Power MOSFET M7 & M14 from Battery to
e-vehicle charging are being used for developing the feature
2.4 Battery Charging Circuit:
of BUCK topology as shown in Fig..2. The e-vehicle
There are two Power MOSFET IRFZ44N M1 & M2 are charging requires always constant current operation because
being used to develop the feature of BUCK Converter from small change in voltage leads to drastic change in current.
Solar Panel to Battery as shown in Fig.3. The MOSFET Gate U27, U26 & U25 are part of driver circuit which is driven by
to Source Voltage is 20 V, the drain current Id is 49A and PWM signal generated by Microcontroller. we get high
RDSON is 17.5 m-Ohm [16]. Based on these basic features power e-vehicle charging range varying from 10 W to 100
further driver circuit designed. [1][2]. the battery voltage is W. For 10 W of e-vehicle charging requires typical drive of
based on the duty cycle D as given in below expression. 3.4V at 350 mA to generate full charging. The maximum
current is 500 mA for the driver circuit. This is the limit for
the circuit; even if it is increasing for short pulse duration

365
 INTERNATIONAL JOURNAL of RENEWABLE ENERGY RESEARCH
A. Patidar et al., Vol.14, No.2, June, 2024

may result in damage of circuit. For 30 W module, similar

circuit can be used but specification will be different i.e. 700
ma for full brightness output and 1A for maximum current
drive for the circuit. [3]
START

3. Control Algorithm Implementation Start Absorption mode,

Controller set
Scan all the parameters To trickle charging
of the controller Check Vpv<14V
Set charging current
The control algorithm utilized in this system primarily Ic=0.20Impp

consists of a Maximum Power Point Tracking (MPPT) Yes

No Yes No
algorithm. However, the circuit features can be divided into
two parts: Check the
STOP Charge Controller
Check
Condition Vpv > Vbat Vbat >13.9V
• Battery Charging Algorithm & Vbat<10.8V
Ipv = 0.8Impp
Reduce charging
current

• Artificial Neural Network Algorithm Yes Yes Yes Stop

Artificial Neural Network Algorithm (ANN): The controller Start Float Charging
Determine the SOC
incorporates a P&O algorithm in each mode, aiming to Of the battery,
Start BULK charging
Mode, set controller
0.20Impp>Ic>0.10Impp

maximize power generation for battery charging [4]. mode

No
Different set points are implemented in the controller, No Yes
enabling all three modes depicted in Figure 6, and facilitating check
12.8V<Vbat<13.9V
the operation of the Perturb and Observe Algorithm. The Check
Vpv>14V

Vbat>12.8V
duty cycle is adjusted in each mode to execute the algorithm.
The flowchart for the Perturb and Observe Algorithm is
illustrated in Fig.3. The duty cycle is varied to increase or
decrease the output voltage of the converter. During each Fig. 3. Battery charging Algorithm
iteration, the algorithm examines the current power state
(P(t)) and voltage state (V(t)), as well as the previous power 4. Result and Discussion
state (P(t-1)) and voltage state (V(t-1)). Based on these
values, the algorithm determines the necessary adjustment to Figure.3 presents the final result output of the battery
the reference duty cycle in order to track the Maximum charging and discharging profile. The graph illustrates the
Power Point. State of Charge (SOC), which increases from 78.42% to
3.1 Battery Charging Algorithm: 78.47% during the battery charging process. Simultaneously,
the battery current increases from 0 Amps to 3.2 Amps. The
It is divided into three modes, as depicted in Figure 3: negative values shown in the battery charging graph indicate
1. Bulk Charging Mode: In this mode, the controller senses that the battery is in a charging state. When the battery
the battery voltage, which typically falls between 10.8V supplies current to the load, the current value becomes
and 12.8V. The controller charges the battery up to 80% of positive.
the short-circuit (SC) current. The charging voltage also increases from 13.3V to 13.5V.
2. Absorption Mode: When the battery is already charged Since the battery is fully charged, the controller operates in
above its nominal voltage of 12.8V, the charge controller the Float charging state, maintaining a constant voltage level.
switches to the constant current charging mode. The
voltage level is maintained until the battery is fully
charged. At this level, the battery can be used to power
loads, but during daytime when lighting loads are
unnecessary, the LiFePO4 battery can be charged up to its
over-voltage level of 14.2V.
3. Float Charging Mode: The third mode of the algorithm is
the float charging mode. In this mode, the battery charge is
maintained between 13.5V and 14.2V. This condition is
considered a floating point condition, where the battery
consumes a very low amperage.
These three modes of operation allow for efficient charging
and maintenance of the LiFePO4 battery, ensuring optimal
utilization and longevity.

Fig. 4. Simulation Result of the MPPT tracking

366
INTERNATIONAL JOURNAL of RENEWABLE ENERGY RESEARCH
A. Patidar et al., Vol.14, No.2, June, 2024

Therefore, the average current during the 12-hour nighttime

period is 0.75A.
Under this load condition, the battery discharges from 13.7V
to 12V over a period of approximately 35 to 40 hours. Since
the battery's nominal voltage is 12.8V, this performance
aligns with the controller's requirements. It is important to
note that the battery's deep discharge voltage is set at 10.8V,
and the system is designed for a 3-day autonomy, ensuring
that it never reaches the deep discharge condition during the
initial stages.
In Fig.6 of the simulation, the battery is observed to
discharge from 13.7V to 13.35V between 23 seconds and 36
seconds. This discharge duration is equivalent to
approximately 12 hours of discharge condition in the
physical battery. This information provides insights into the
discharge behavior of the battery, ensuring it operates within
Fig. 5. Simulation Result of the system performance with
the desired voltage range throughout the nighttime period.
various features
Table 1. Experimental Data Recorded
If we examine the MPPT power conversion, we can observe
that the maximum power delivered by the solar panel SOC Charging Voltage Discharge Voltage
(represented by the blue line) fluctuates between 55W and
58W. However, the maximum power consumed by the (in %) (in Volts) (in Volts)
battery ranges from 45W to 50W. This difference is due to 0
the fact that when the battery is in a fully charged condition,
it requires minimal charging, resulting in lower power 5
consumption.
10 12.8
Overall, the graph provides insights into the battery charging
process, including SOC, current, voltage, and power 15 12.95 12.85
consumption, highlighting the efficient operation of the 20 13.04 12.9
system.
25 13.13 12.98
When examining the battery discharging characteristic, we
plotted the data for various load conditions. The battery was 30 13.2 13.02
observed to deliver currents ranging from 0.5A to 20A, but 35 13.23 13.07
our primary interest lies in the discharge characteristic at
0.75A. The physical controller is designed to allow a 40 13.25 13.15
maximum of 1A during normal night hours and 0.5A during
45 13.25 13.17
the dimming mode after 6 hours of initial night hours.
50 13.26 13.18
55 13.26 13.19
60 13.27 13.195
65 13.275 13.2
70 13.3 13.22
75 13.35 13.27
80 13.38 13.3
85 13.4 13.34
90 13.41 13.35
95 13.43
100 13.48

Fig. 6. Battery Discharge Characteristic as per simulation

result

367
INTERNATIONAL JOURNAL of RENEWABLE ENERGY RESEARCH
A. Patidar et al., Vol.14, No.2, June, 2024

In Table 1, the actual data recorded from hardware design, policies by accelerating EV adoption, spurring innovation,
showing only nominal change in discharge and charge cycle and driving sustainable energy practices.
of the battery. The same effect has mentioned in Fig.7, where
it clearly showing that the charging and discharging cycle are
not in similar manner.
References
1. R. M. Elavarasan, G.M. Shafiullah, S.K. Padmanaban,
Nallapaneni M.K., A. Annam, A.M. Vetrichelvan, L
M Popa, J.B. Holm-Nelson, “A Comprehensive
Review on Renewable Energy Development,
Challenges, and Policies of Leading Indian States
With an International Perspective”, IEEE Access, Vol.
8, pp. 74432-74457, Apr. 2020.
2. M. Adley, Kai Strunz, “Irradiance-Adaptive PV
Module Integrated Converter for High Efficiency and
Power Quality in Standalone and DC Microgrid
Applications”, IEEE Transaction on Industrial
Electronics,vol. 65, no. 1, pp. 436-446, Jan-2018.
Figure 7. Battery charging and discharging curve with
respect to the SOC 3. V. K. Dubey and S. K. Sharma, "Efficient Driver
Circuit Design ameliorating MPPT Tracking at Solar
5. Conclusion based Lithium Ion Battery Charging," 2022 IEEE
Delhi Section Conference (DELCON), New Delhi,
The system becomes very much efficient if we realize all India, 2022, pp. 1-8, doi:
the performance parameter in virtual simulation. Sim- 10.1109/DELCON54057.2022.9753385.
electronics in MATLAB is very much advance library where
we can work on real time system. As I did the simulation in 4. K. Nunotani, F. Yoshida, Y. Kamiya, Y. Daisho, K.
MATLAB, I got the same result in hardware, due to which I Abe, M. Kono, H. Matsuo, “Development and
got the same efficiency in P&O algorithm as you can get in Performance evaluation of Lithium ion Phosphate
INC algorithm. Therefore this topology for the buck Battery with superior rapid changing performance”,
converter has played the significant role to enhance the 2011 IEEE Vehicle Power and Propulsion Conference,
efficiency of the system. The SOC estimation of Lithium Ion Chicago, USA, Sept 2011.
batteries is very difficult. The typical discharge curve of the 5. J.T.D.C. Neto, A.O. Salazar, A. S. Lock, “Design and
Lithium Ferrous Phosphate battery, Lithium Manganese Operation of an OCC-based scheme for Stand-Alone
Oxide Battery and Lead Acid battery on same voltage level PV system powering the DC Load”, IET Renewable
with respect to the DOD of the battery have been mentioned Power Generation, Vol 13, no. 12, Sept 2019.
in figure The diagram shows that voltage measurement from
40% to 80% is 6% while in case of LFP batteries, it is only 6. J.J. Soon, K. Low, “Optimizing Photovoltaic Model
0.5 V variations. The similar variations have been found in for Different Cell Technologies Using A Generalized
our observation. Lithium Ferrous Phosphate batteries has the Multi-dimension Diode Model”, IEEE Transactions
advantage of very high level of safety compare to other on Industrial Electronics, Vol. 62, no. 10, pp. 6371-
battery technologies. Due to its High Power density, it is 6380, Oct. 2015.
used in various applications like as medium power traction 7. K.J. Sauer, T. Roessler, C.W. Hansen, “Modeling the
for example robotics, AGV, E-mobility. This is only due to Irradiance and Temperature Dependence of
its characteristics. As increasing demand of Stand-Alone Photovoltaic Modules in PVsyst”, IEEE Journal of
Solar system to meet the power requirement has led to the Photovoltaics, pp. 152-158, Nov. 2014.
revolution in non-conventional energy source technology.
8. F. Reis, C. Guerreiro, F. Batista, T. Pimental, M.
As it improved the charging experience for EV users, may Pravettoni, J. Wemnes, G. Sorasio, M.C. Brito,
increase convenience and reduced charging time can “Modeling the Effects of Inhomogeneous Irradiation
incentivize more individuals to switch to electric vehicles. and Temperature Profile on CPV Solar Cell
Behavior”, IEEE Journal of Photovoltaics, Vol. 5, no.
The implementation of fast chargers utilizing MPPT 1, pp. 112-122, Jan. 2015.
algorithms has the potential to reshape the Indian energy 9. E.I. Batzelis, S.A. Papathanassiou, “A Method for the
sector's landscape. This transition would necessitate strategic Analytical Extraction of the Single-Diode PV Model
planning, investments in infrastructure, grid modernization, Parameters”, IEEE Transactions on Sustainable
and policy adaptations to ensure a seamless integration of Energy, Vol.7, no.2, pp. 504-512, Dec. 2015.
EVs into the transportation network. The changes brought
about by this technology could play a crucial role in 10. I.A. Ibrahim, Benjamin C. Duck, Christopher J. Fell,
promoting a progressive environment for EV manufacturing “An Adaptive Wind Driven Optimization Algorithm
for Extracting the Parameters of Single Diode PV cell

368
INTERNATIONAL JOURNAL of RENEWABLE ENERGY RESEARCH
A. Patidar et al., Vol.14, No.2, June, 2024

model”, IEEE Transaction on Sustainable Energy, vol.

11, no 2, pp 1054-1066, Apr 2020.
11. G. R. C. mouli, J. Schijffellen, M.V.D.Heuvel, M.
Kardolus, P. Bauer, “A 10kW Solar-Powered
Bidirectional EV Charger Compatible with Chademo
and COMBO”, IEEE Transactions on Power
Electronics, vol. 34, no. 2, pp. 1082-1098, Feb. 2019.
12. Andrea Montecucco, Andrew R. Knox, “Maximum
Power Point Tracking Converter Based on the Open-
Circuit Voltage Method for Thermoelectric
Generators”, IEEE Transaction on Power Electronics,
vol 30, no 2, pp 828-838, Feb 2015.
13. N. MOHAN, T. UNDELAND, and W. ROBBINS,
Power Electronics: Converters, Applications, and
Design, 2nd edit., New York: John Wiley & Sons,
1995, Chapters 19-26.
14. R. W. Erickson, Dragan Maksimovic, Fundamentals
of Power Electronics, 2nd Edit., Bouldor, Colarado:
New Yiork: Cluver Academic Publisher, 2001,
Chapter 1-10.
15. S. Teleke, M.E.Baran, S. Bhattacharya, A.Q. Huang,
“Rule-Based Control of Battery Energy Storage for
Dispatching Intermittent Renewable Sources”, IEEE
Transactions on Sustainable Energy, vol.1, no.3,
pp.117-124, Oct. 2010.
16. Al Khas, A. and Cicek, I., 2019, November. SHA-512
based wireless authentication scheme for smart battery
management systems. In 2019 8th International
Conference on Renewable Energy Research and
Applications (ICRERA) (pp. 968-972). IEEE.
17. Harighi, T., Bayindir, R. and Hossain, E., 2018.
Overviewing quality of electric vehicle charging
stations’ service evaluation. International Journal of
Smart Grid-ijSmartGrid, 2(1), pp.40-48.

369