Ain Shams Engineering Journal xxx (xxxx) xxx

Contents lists available at ScienceDirect

Ain Shams Engineering Journal

journal homepage: www.sciencedirect.com

Nonlinear coordination strategy between renewable energy sources and

fuel cells for frequency regulation of hybrid power systems
Fahad M. Almasoudi a, Abualkasim Bakeer b, Gaber Magdy c, d, Khaled Saleem S. Alatawi a, *,
Gaber Shabib c, e, Abderrahim Lakhouit f, Sultan E. Alomrani a
a
Electrical Engineering Department, Faculty of Engineering, University of Tabuk, Tabuk 47913, Saudi Arabia
b
Department of Electrical Engineering, Faculty of Engineering, Aswan University, Aswan 81542, Egypt
c
Department of Electrical Engineering, Faculty of Energy Engineering, Aswan University, Aswan 81528, Egypt
d
Faculty of Engineering, King Salman International University, El-Tor, South Sinai 46511, Egypt
e
Higher Institute of Computer, King Marriott, Alexandria 23713, Egypt
f
Department of Civil Engineering, Faculty of Engineering, University of Tabuk, Tabuk 47913, Saudi Arabia

A R T I C L E I N F O A B S T R A C T

Keywords: This study proposes an advanced control strategy for the coordination of an energy storage system (ESS) based on
Renewable energy sources (RESs) fuel cells (FCs) and renewable energy sources (RESs) to enhance frequency dynamic performance in hybrid
Fuel cell (FC) power systems (HPSs). The proposed coordination control strategy is based on the nonlinear proportional-
Nonlinear PI (NPI) controller
integral (NPI) controller, which increases the system’s flexibility in dealing with disturbances and changing
Dandelion optimizer (DO)
Load frequency control (LFC)
operating conditions. In addition, it improves the system’s dynamic response and attempts to address system
Hybrid power system (HPS) weakness caused by highly penetrating RESs. The proposed NPI controller is optimally designed using a new
optimization algorithm, called dandelion optimizer (DO), whose proficiency and effectiveness are verified by
comparing its performance with other well-known optimization algorithms used in the literature; particle swarm
optimization (PSO), grey wolf optimization (GWO), and ant lion optimization (ALO) algorithms considering
various standard objective functions. Furthermore, the proposed NPI controller performs better than other
control strategies used in the literature under load/RESs fluctuations. The effectiveness of the proposed nonlinear
coordination control strategy is examined and investigated through a self-contained HPS that includes a diesel
generator, RESs (i.e., photovoltaic and wind power plants), battery ESS, flywheel ESS, aqua electrolyzer for
hydrogen production, FCs, electric vehicles, and customer loads. The simulation results carried out by the
MATLAB software demonstrate the superior performance of the proposed DO-optimized NPI controller for HPS
frequency regulation, even when the power system’s parameters have substantial variations. Moreover, the re­
sults revealed that the proposed strategy significantly reduces the frequency deviation by approximately 95%
compared to the conventional coordination strategy based on the fixed contribution of RESs and by 90%
compared to the adaptive coordination control based on the PI controller.

1. Introduction Status Report” [1], 93 GW worth of electrical wind energy was added to
the grid, increasing the system’s overall capacity to 743 GW. Although
Most world countries face the key challenge of upgrading the power 193 GW of solar photovoltaic (PV) entered service globally, increasing
grid infrastructure through large-scale renewable energy sources (RESs) the total renewable solar photovoltaic capacity to 760 GW. On the other
integration, as environmentally friendly programs for renewable energy hand, the unpredictable nature of weather conditions results in the
are on many government agendas. In addition, the need to minimize intermittent generation of RESs, as these sources depend on climatic
environmental consequences and economic growth and improve the factors, including the amount of solar irradiation and the wind velocity
resistance of the electrical power grid to disruptions has resulted in an [2]. Thus, oscillations in system voltage/frequency result from the
increase in the use of RESs and the transition from centralized to unpredictability of RESs generation, as well as the probabilistic behavior
distributed generation. According to the “Renewables 2021 Global of the loads [3–9]. Depending on demand variations, energy storage

* Corresponding author.
E-mail address: khaled@ut.edu.sa (K.S.S. Alatawi).

https://doi.org/10.1016/j.asej.2023.102399
Received 9 February 2023; Received in revised form 17 June 2023; Accepted 17 July 2023
2090-4479/© 2023 THE AUTHORS. Published by Elsevier BV on behalf of Faculty of Engineering, Ain Shams University. This is an open access article under the CC
BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Please cite this article as: Fahad M. Almasoudi et al., Ain Shams Engineering Journal, https://doi.org/10.1016/j.asej.2023.102399
F.M. Almasoudi et al. Ain Shams Engineering Journal xxx (xxxx) xxx

Fig. 1. The dynamic model of the HPS with the proposed adaptive coordination strategy based on NPI.

systems (ESSs) such as flywheels, batteries, ultracapacitors, and super­ based on power and voltage control has been proposed for a hybrid ac/
conducting magnetic energy storage store surplus power and release dc microgrid that includes photovoltaic, wind, and battery [15].
deficient power to the grid [10]. Furthermore, an effective coordination strategy between ESS based on
A hybrid power system (HPS) is considered a strategic solution to FC and RESs is being suggested as a means of improving the stability of
face the key challenges of integrating RESs into the power grid, where system frequency of HPS along with numerous load frequency control
HPSs are a combination of multi-electrical power sources, and their (LFC) techniques, such as a proportional integral derivative (PID)
operation is coordinated utilizing power electronics systems [11]. controller based on a modified moth swarm algorithm [16], fractional-
Moreover, the HPS has RESs alone or in conjunction with conventional order control technique [17], and cascaded fuzzy-PID control strategy
energy sources. The main feature of HPS is that it is self-sufficient in a [18]. The electrolysis process is powered by the excess energy that the
wide range of climatic conditions and does not rely on a single source. It RES generates through an aqua electrolyzer to store hydrogen energy in
can be run independently as a microgrid or incorporated into the power FCs. These studies considered this excess power from RESs a constant
grid. Energy storage systems (ESSs) may be integrated with HPSs to fraction of total renewable generation (Kn). However, to achieve the best
mitigate the negative impacts of RESs since their power is inherently possible economic benefit from RESs, the portion of total power gener­
random and depends on environmental conditions. This may lead to ation from RESs used for hydrogen energy storage in FCs should vary
instances where the electrical load exceeds the generation [12]. These with system conditions. Therefore, to address this drawback, adaptive
ESSs also reduce grid frequency fluctuations and improve power quality. coordination between ESS based on FC and RESs has been proposed
If excess power is available from RESs on the load demand, these ESSs based on a PI controller for frequency stabilization of HPS [12].
will store it briefly before transferring it to the main grid when the load On the other side, several control techniques have been applied for
demands are more significant than the power production. As discussed the LFC of power systems considering RESs and ESSs, such as; the PID
in this research, a controlling mechanism that coordinates numerous controller [19,20,21], fractional-order PID controller [17,22], fuzzy
operations is required for these operations to be carried out logic control [23], model predictive control [24], H_infinty technique
appropriately. [25]. Moreover, a modified African Buffalo Optimization algorithm has
Numerous coordination control schemes for RESs and ESSs for HPSs been applied to design several frequency controllers for a multi-
have been implemented in the literature. The authors in [13] developed microgrid system with heterogeneous generation sources [26].
an HPS including a wind generator, PV, electrolyzation, and fuel cell Furthermore, An adaptive LFC based on a modified Jaya optimization
(FC) and presented the coordination control simulation studies. algorithm has been applied to enhance the frequency stability of the
Furthermore, an optimally coordinated scheme has been proposed for islanded/interconnected power systems [27]. An efficient cascaded
wind, photovoltaic, hydrogen storage, and FC units in a hybrid distrib­ control strategy has been proposed for frequency regulation of two-area
uted power system [14]. An effective coordination control strategy HPS considering electric vehicles [28]. A discrete-time LFC model based

2
F.M. Almasoudi et al. Ain Shams Engineering Journal xxx (xxxx) xxx

Table 1 2. Hybrid power system (HPS) model

HPS analyzed parameters and their values.
Components Gain Time constant The HPS proposed in this study contains various types of energy
resources: RESs (PV and wind systems), an aqua electrolyzer (AE), fuel
WG KWG = 1.0 TWG = 1.5
PV KPV = 1.0 TPV = 1.8 cells (FCs), electric vehicles (EVs), flywheel energy storage system
FC KFC = 0.01 TFC = 4.0 (FESS), battery energy storage system (BESS), and a diesel generator
DG KDG = 0.003 TDG = 2.0 (DG). A similar system has been extensively studied in [11]. The sche­
AE KAE = 0.002 TAE = 0.5 matic and transfer functions modeling of the proposed HPS is presented
EV KEV = 1.0 TEV = 1.0
BESS KBESS = − 0.003 TBESS = 0.1
in Fig. 1. Each component model’s gain and time constant are listed in
FESS KFESS = − 0.01 TFESS = 0.1 Table 1.
Power system
Parameter Damping constant Inertia constant A. Generation Subsystem Models
Value D = 0.3 J = 0.4

The WG, PV, FC, and DG are modelled by first-order transfer func­
on Tustin’s technique has been proposed for a real multi-source power tions in (1)-(4) with the appropriate time constant and gain, as shown in
system [29]. Table 1.
In light of the aforementioned reasons, effective methods to regulate KWG ΔPWG
the extra energy utilized from RESs to energize the ESS based on FC have GWG (s) = = (1)
1 + sT WG ΔPm
become highly critical to cope with loads/RESs uncertainty concerns
present in today’s HPS. It is often possible to control a physical process KPV ΔPPV
GPV (s) = = (2)
using PI controllers with a linear fixed gain. However, in many cases, 1 + sTPV Δϕ
high-performance control requires greater flexibility to cope with
changes in operating or environmental conditions [30]. Therefore, GFC (s) =
KFC
=
ΔPFC
(3)
several approaches have been advanced to enhance the resilience and 1 + sTFC ΔPAE
robustness of linear PI controllers, including fuzzy logic control, self-
tuning method, predictive control, and others. Among these ap­ KDG ΔPDG
GDG (s) = = (4)
proaches, a nonlinear PI (NPI) controller is regarded as one of the most 1 + sTDG ΔuDG
effective and simple methods for industrial applications [31–33].
Therefore, the NPI controller increases the system’s flexibility in dealing B. Aqua Electrolyzer Model
with disturbances and changing operating conditions. Moreover, it im­
proves the HPS’s dynamic response and attempts to treat the system’s The AE uses the surplus power generated by RESs to produce
weakening caused by high-penetration RESs. hydrogen, which will be injected into FCs to generate power. The first-
The following are the main contributions to this paper: order transfer function models the AE as [11]:
KAE ΔPAE
I. Propose a sophisticated control strategy based on an NPI GAE (s) = = (5)
controller to coordinate ESS based on FC and RESs to enhance 1 + sTAE (1 − Kn ) × PRESs
frequency dynamic performance in HPS considering large-scale Kn and ΔPAE represent the fraction of total renewable generation that
RESs. Moreover, the superior performance of the proposed NPI energies the AE and the output of the aqua electrolyzer, respectively.
controller is confirmed by comparing it with various controlling
techniques discussed in the published works, such as in [12,16] C. Electric Vehicles Model
under load/RESs fluctuations.
II. The suggested NPI controller uses an innovative optimization The EVs performance helps to mitigate any fluctuations in the system
method known as the dandelion optimizer (DO) to achieve opti­ because they can function both as a load when the system is in the
mum design, which is applied for the first time in the LFC issue. In charging mode and as a supply when it is in the discharging mode.
addition to this, the effectiveness of the suggested DO algorithm The first-order transfer function models the EVs system as [11]:
is validated by comparison to other effective optimization stra­
KEV
tegies found in the published research; particle swarm optimi­ GEV (s) = (6)
zation (PSO), grey wolf optimization (GWO), and ant lion 1 + sTEV
optimization (ALO) algorithms taking into account various
standard objective functions. D. Energy Storage Systems Model
III. The proposed NPI controllers’ design procedures take into ac­
count the uncertainties of RESs/loads. The ESSs are crucial to maintaining the system frequency since they
IV. The validity of the suggested coordination controlling technique can work as a load or a power supply. The BESS and FESS are modelled
is examined and investigated through a self-contained HPS that by the first-order transfer functions (7)–(8).
includes RESs (PV and wind systems), FCs, a diesel generator
KBESS
(DG), electric vehicles (EVs), battery ESS (BESS), flywheel ESS GBESS (s) = (7)
1 + sTBESS
(FESS), and customer loads.
KFESS
The rest of this paper is organized as follows: The studied HPS GFESS (s) = (8)
1 + sTFESS
configuration and modeling are presented in Section 2. Section 3 pro­
vides the proposed NPI controller and the optimized algorithm based on
the DO. The results of the simulations carried out using the MATLAB/ E. Power System Model
Simulink program are presented in Section 4. In the last part of the work,
the conclusion is summarized in Section 5. In HPS, based on the difference between the generating power and
the load demand, the frequency fluctuates. The frequency deviation (Δf )
increases with increasing power deviation (ΔP) and vice versa. As a

3
F.M. Almasoudi et al. Ain Shams Engineering Journal xxx (xxxx) xxx

linear PI controllers may be a sufficient method for controlling physical

processes without changing operating conditions or system parameters.
However, high-performance control is often beyond the capabilities of
simple PI controllers. In that case, the nonlinear PI (NPI) controller
becomes a useful solution for many applications to improve the system’s
performance [30]. In addition, the NPI controller allows for more flex­
ibility, can handle wide-range disturbances, and increases tracking ac­
curacy compared to the PI controller. The output of the NPI controller in
the time domain has the given form in Eq. (12), while the main structure
of the NPI controller is sketched in Fig. 2. With NPI controllers,
nonlinear gains are combined with fixed gains [31]. Furthermore, unlike
the conventional PI controller, which has only two parameters to be
tuned, the NPI controller has three parameters that can improve tran­
sient time, stability, and steady-state error. As a result, the overall NPI
controller can improve the steady-state and dynamic performance of the
controlled system.
( ∫t )
u(t) = f (e) Kp e(t) + Ki e(t)dt (10)
Fig. 2. The basic structure of the NPI controller. 0

result, the power deviation determines the frequency deviation as where Kp and Ki are the proportional and integral gains of the NPI
controller, respectively, e(t) is the error, and f(e) is the nonlinear gain
1 Δf
GPS (s) = = (9) function.
D + sJ ΔP The nonlinear gain function f(e) has many forms and can be imple­
where D is the HPS’s damping constant and the equivalent mass of the mented in cascade with a linear static gain of the PI controller. In this
rotating parts in the HPS is represented by the inertia values J. paper, the utilized form of f(e) is given in Eq. (13), in which k is a
positive gain to be tuned [32]. As f(e) is a function of the error, the
3. Proposed nonlinear PI controller nonlinear gain is initially very high and then becomes low to avoid
unstable behavior [33]. The high initial value ensures a fast dynamic
With increasing the complexity and nonlinearity of the system, the response, and the low gain allows it to remain stable.
conventional controllers based on the PI with fixed parameters become eke(t) + e− ke(t)

insufficient to achieve the desired performance of the system. Fixed-gain f (e) = (11)
2

Fig. 3. Proposed adaptive coordination strategy based on NPI with HPS to optimally distribute energy among RESs and FC-ESS.

4
F.M. Almasoudi et al. Ain Shams Engineering Journal xxx (xxxx) xxx

Fig. 4. The tuning process of the NPI gains for optimal coordination between the RESs and FC-ESS in the studied HPS.

The studied HPS is considered to be a hybrid generation system

containing different distributed energy sources of RESs, ESSs, and diesel
generators. As these RESs have a high penetration level in the studied
power system, the system’s frequency will become more prone to fluc­
tuations where the RESs do not have natural inertia properties.
Furthermore, the uncertainty of PV and wind systems and load proba­
bilities lead to system frequency fluctuations. Therefore, the proposed
adaptive energy coordination will be applied to coordinate between
these distributed generations (DGs) sources to stabilize the power sys­
tem’s frequency (i.e., more specifically, between RESs and FC-ESS). The
proposed coordination control strategy consists of deriving electrical
power-yielding capacity from the RESs to hydrogen energy by a
hydrogen processing system based on the FC-ESS, as shown in Fig. 3.
Similarly, the control input to the other DGs in the HPS will be set by the
conventional LFC based on mMSA, as described in [16].
Similarly, the power produced by the other DGs is solely affected by
the deviation in the system frequency. Among those functions, the
implementation of adaptive control methods for the RESs, AEs to pro­
duce hydrogen, and FC-ESS will be implemented, as well as the
improvement of the frequency control of the modern power system in
addition to its main frequency control. Moreover, the proposed coordi­
nation strategy is adaptive to quickly mitigate the system disturbances
where the RESs power is considered a disturbance source. A coordina­
tion gain of one means that the FC-ESS will not deliver the HPS, as the
HPS does not require extra power, and the FC-ESS can store all the
available RESs energy to be used if a disturbance occurs.
There is a trial-and-error methodology for defining the gains of the
NPI controller (i.e., Kp, Ki, and k). It makes the process of tuning very
challenging since it is dependent on the expertise of the controller
designer. In addition, using these approaches, which involve a process of
trial and error, makes it challenging to determine the ideal parameter
value for the NPI gains, which may boost the system’s overall perfor­
mance and ensure the system’s resilience to disruptions. As a direct
result, there is no assurance that the system will maintain its resilience in
the given circumstances. To cope with this problem, in this research, the
optimal values for NPI controller parameters are determined through a
metaheuristic optimization approach based on the DO algorithm, aiming
to reduce frequency deviations and improve power system stability. The
obtained optimal NPI gains can be used to calculate the excess energy
from RESs (i.e., the percentage of RES power generation, Kn) and deliver
it into the FC-ESS. The overall tuning process of the NPI gains is illus­
trated in Fig. 4.Fig. 5
In this study, the DO is applied in the MATLAB program. The
employed optimization method plays a significant role in achieving the
adaptive control method of the RESs and the aqua electrolyzer. In
addition to enhancing the frequency of the main control, the optimiza­
tion method enhances the frequency of the modern power system under
Fig. 5. The flowchart of the DO algorithm.
different load conditions. The objective functions under consideration
are the integral of time multiplied absolute error (ITAE), the integral of
absolute error (IAE), the integral of time multiplied squared error (ITSE),

5
F.M. Almasoudi et al. Ain Shams Engineering Journal xxx (xxxx) xxx

and the integral of squared error (ISE), and their formulae are provided evolve their population.
in Eqs. (14) to (17). The error in this case study is the system frequency
deviation (Δf). I Rising Stage
∫ tsim
ITAE = (|Δf |)*tdt (12) Before dandelion seeds become independent of their parent plant,
0 they must develop to a specific maturity level and reach a certain height,
∫ tsim depending on weather conditions. The main two meteorological con­
IAE = (|Δf |)dt (13) ditions in this case are:
0
Condition (1): Wind speed might be considered to have a log-normal
∫ tsim distribution.
( )
ITSE = (Δf 2 )*tdt (14) lnY N μσ2 on a clear day. The random values are more uniformly
dispersed along the Y-axis thanks to this distribution, so there is a
0

∫ tsim greater chance that dandelion seeds will disperse to distant areas. Thus,
ISE = (Δf 2 )dt (15) DO places a strong emphasis on exploration during Case 1. Dandelion
seeds are dispersed at random by the wind in the search space. The wind
0

where tsim is the simulation time. speed determines the height at which a dandelion seed will rise.
Dandelion seeds disperse farther and higher as the wind strength in­
B. Optimization Approach Based on DO creases. The vortices that develop above the dandelion seeds are
constantly being reshaped as a direct consequence of variations in wind
Dandelion optimization (DO) is one of the recent optimization speed. This causes the vortices to ascend in a spiral configuration, which
techniques; it amplifies the movement patterns of dandelion seeds as may be represented as follows:
they fly, which, in comparison to other metaheuristic algorithms, is Pt+1 = Pt + α*vx *vy *lnY*(Ps − Ps ) (16)
capable of reaching convergence with a higher degree of precision using
a less number of iterations. A dandelion is a perennial herb that belongs where Pt is the dandelion seed position during t iteration. Ps is the
to the Asteraceae family, and its scientific name is Herba taraxc. These location in the search space that was randomly produced during the t
plants have a maximum height of more than 20 cm. The heads of the iteration, and it may be written as:
dandelion resemble inflorescences. The seeds typically consist of
achene, a beak, and hundreds of crested hairs [34]. The wind transports Ps = rand(1.Dim)*(UB − LB) + LB (17)
the mature seeds to new areas where they can reproduce life. Crown
hairs are crucial for the movement of Asteraceae seedlings. These hairs lognormal distributioniIs lnY where(μ= 0), (σ 2 = 1), which can be
slow the seeds’ descent, which allows the wind to carry the seeds farther expressed as:
⎧ [ ]
[34]. The common dandelion is one plant that uses the wind to spread its
⎨ √1̅̅̅̅̅ exp − 1 (lny)2
⎪
seeds. Under the correct conditions, its seed can travel for dozens of lnY = y 2π 2σ 2 y≥0
(18)
kilometers. A dandelion seed creates two vortices as it flies, which ⎪
⎩ y<0
0
causes upward drag. The two vortices above expand and align sym­
metrically when the seeds fall at slower speeds [34]. An asymmetric
where the standard normal distribution is y. The adaptive parameter
vortex, in which the fruit points downwards and the filament is parallel
that determines how large or small the search steps should be α which
to the ground, ensures the smooth descent of the seed. Dandelion seeds
can be expressed as:
must be kept at a somewhat constant altitude to travel great distances.
( )
Below the dandelion crown, the divided vortex is kept at a constant 1 2
α = rand( )* 2 t2 − t + 1 (19)
distance. Intriguingly, the dandelion crown porosity seems perfectly T T
controlled to maintain the vortex ring. The thin filaments that make up
vx and vy is the dandelion lift coefficients caused by separated eddy
the crowned hairs branch out from a central handle in a pattern like the
action and can be calculated by:
spokes on a bicycle wheel. These seeds consistently have 90–110 spokes.
The separation vortex above a dandelion seed is steady because of this vx = r*cosθ (20)
uniformity, which helps the seed maintain its stability during long-
distance flight [35]. The two main parameters influencing the vy = r*sinθ (21)
dispersal of dandelion seeds are wind speed and weather. A seed will fly
far or close to home depending on the wind speed. The weather in­ where θ is a number chosen randomly between [ − π, π ] and r = e1θ .
fluences the ability of dandelion seeds to fly and the likelihood that they Condition (2): On a rainy day, humidity, air resistance, and other
will spread to neighboring or distant areas. The dandelion seeds undergo factors prevent dandelion seeds from rising properly. Therefore, the
three stages as follows: seeds of dandelion plants are exploited locally. This case can be
expressed as:
• When a dandelion seed is at the stage when it is ascending, a vortex is
Pt+1 = Pt *K (22)
produced above it, and as a consequence of a pulling force under
circumstances where there is wind and sunshine, the seed rises. In where K is a dandelion local domain search regulator coefficient, and it
contrast, in rainy weather, no eddies are above the seeds. Therefore, can be calculated by:
local searches are the only options in this situation.
• In the descending stage, the seeds start to drop gradually once they K = 1 − rand( )*q (23)
reach a certain height. ( )
• In the landing stage, under wind influence, the seed will land q=
1
t2 − 2
2
t+1+ 2
1
(24)
randomly in a location to reproduce new dandelions. T 2 − 2T + 1 T − 2T + 1 T − 2T + 1
In conclusion, the growing phase of dandelion seeds may be
The DO algorithm was primarily inspired by the three stages in described as follows:
which dandelions disperse their seeds to the subsequent generation to

6
F.M. Almasoudi et al. Ain Shams Engineering Journal xxx (xxxx) xxx

Fig. 6. Convergence curve of different optimization approaches for tuning the proposed NPI parameters considering different formulas for the objective functions.

{
Pt + α*vx *vy *lnY*(Ps − Ps ) randn < 1.5
Table 2 Pt+1 = (25)
Pt *K else
Summary of the comparison between the studied metaheuristics optimization
approaches at different formulas of the objective functions.
where randn is a random number that falls within the normal
Objective function Optimization Algorithm distribution.
DO (Proposed) GWO ALO PSO

ISE 0.10124 0.10968 0.29019 0.1019 ii Descending stage

ITSE 2.4744 3.0235 2.5319 2.5896
IAE 1.1446 1.1914 2.0085 1.1676 After reaching a particular altitude, dandelion seeds begin their slow
ITAE 5.3091 28.484 40.5146 28.5950
and steady descent to the ground. In this technique, the movement
trajectory of dandelions is modeled using brown motion, which makes it
simpler for people to traverse through more search communities
throughout each iteration. This stage can be expressed as:

Fig. 7. Dynamic response of the power system’s frequency with optimal parameters in different formulas for the proposed DO with a uniform change in RESs profile.

Fig. 8. Dynamic response of the power system’s frequency with the optimal parameters in different formulas for the proposed DO with a random profile of RESs.

7
F.M. Almasoudi et al. Ain Shams Engineering Journal xxx (xxxx) xxx

Table 3
Summary of the optimal values of NPI with the proposed DO in different formulas of the objective function.
Objective function Nonlinear PI controller-based DO algorithm parameters Performance indices of system frequency

Kp Ki k MUS (Hz) MOS (Hz) Ts (s)

ISE 0.10124 40.5660 32.1730 0.231019 0.000015 0.007460 4.2490

ITSE 2.4744 39.4678 19.9046 0.144446 0.000083 0.007860 6.6100
IAE 1.1446 39.4628 20.1323 0.957835 0.000010 0.007856 6.7040
ITAE 5.3091 0.87805 2.3987 0.000086 0.007205 0.038400 16.586

Fig. 9. Dynamic response of the power system with the proposed NPI controller compared to literature works in [12,16].

Pt+1 = Pt − α*βt *(Pmean t − α*βt *Pt ) (26)

Pt+1 = Pelite + α*levy(λ)*(Pelite − Pt *δ) (28)
where the Brownian motion is βt that is a randomly chosen number from
where Pelite is the dandelion seed’s optimal position. δ is between [0, 2],
the standard normal distribution. The population’s average position is
and it represents the linear increasing function and can be calculated by:
denoted by Pmean t that can be expressed by:
2t
1 ∑
pop
δ= (29)
Pmean t = Xi (27) T
pop i=1
levy(λ) is the levy flight function as:
w×σ
iii Landing Stage levy(λ) = 0.01 × 1 (30)
|t|β
The dandelion seed makes a haphazard decision on where to settle
where β is a number chosen randomly between [0,2]. wandt are
down, considering both the ascending and descending phases. The ideal
randomly chosen numbers between [0,1] and σ can be calculated by:
answer that will be reached as the algorithm progresses is the general
⎛ ⎞
location where the dandelion seed will likely fall. This will be accom­
plished after the program has completed its progression. This stage may ⎜Γ(1 + β) × sin(π2β)⎟
σ=⎜ ⎟ (31)
be characterized by the following: ⎝
Γ(1+β )×β×2 2
β− 1 ⎠

8
F.M. Almasoudi et al. Ain Shams Engineering Journal xxx (xxxx) xxx

Fig. 10. Dynamic response of the power system with the proposed NPI controller compared to literature works in [12,16].

proposed DO algorithm has the lowest value for the four objective
4. Results and discussion functions compared to the other metaheuristic methods.
The optimal solution for the NPI variables can be determined using
The studied HPS has been simulated based on MATLAB/Simulink the best cost function to regulate the system frequency. However,
using the parameters given in Table 1. First, compared to other meta­ because each cost function has a unique mathematical formula in terms
heuristic algorithms like GWO, PSO, and ALO, we have demonstrated of the number of inputs it depends on and how they are related to one
the viability of the employed optimization approach based on the DO at another, it is impossible to compare the cost functions based only on the
different formulas of the fitness function. Then, the proposed NPI final value acquired during the execution of the algorithm. Therefore,
controller is tested under various loading and RESs scenarios using the we have separately tested the optimal values of each type of fitness
optimal parameters of the objective function that exhibits the best per­ function with the proposed DO algorithm (i.e., listed in Table 2) in the
formance for the frequency deviation in terms of the settling time (Ts), simulation, as shown in Fig. 7 and Fig. 8. In Fig. 7, the load is stepped up
maximum overshoot (MOS), and maximum undershoot (MUS). Based on to 0.1 pu at 50 s. In comparison, the RESs profile is uniform and provides
DO optimization, the proposed NPI controller has been compared with 0.25 pu for a period of 30 s (i.e., from 30 to 60 s). In this case, the fre­
the fixed ratio coordination factor between the RESs and FC-ESS and the quency deviation of the HPS exhibits an overshoot at the instant of 50 s
adaptive coordination gain based on the traditional PI controller. To where the load is changed. While in Fig. 8, the same load profile in
ensure a fair comparison of the controllers studied, we have used the Fig. 7a is applied, but a randomly RESs profile is utilized to test the
same controller of the main frequency regulation based on mMSA in all robustness of the proposed NPI controller with each fitness function
the scenarios studied [16]. Finally, the cumulative objective function of against the RESs uncertainties. The zoom plot in the frequency devia­
the three coordination strategies is evaluated for each scenario. tion, as shown in Fig. 8b, demonstrates that the ISE provides the mini­
mum overshoot in the frequency deviation compared to other fitness
4.1. Effectiveness of the proposed DO algorithm functions.
In addition, Table 3 shows the performance indices associated with
Fig. 6 depicts a comparison of various optimization techniques the system’s frequency with each fitness function, such as Ts, MOS, and
considering various objective functions. The proposed DO algorithm MUS. Despite the MOS and MUS of the ISE and IAE being almost the best
achieves the fastest convergence curve with the smallest value for each among the fitness functions and almost being almost the closer, ISE
type of fitness function. In addition, the DO algorithm nearly finds the improves the settling time of the frequency deviation by approximately
optimal parameters after 30, 25, 18, and 15 iterations for the ISE, ITSE, 35% compared to the IAE. Therefore, the proposed NPI controller-based
IAE, and ITAE, respectively. Finally, Table 2 summarizes the value of DO considers the optimal parameters based on the ISE’s objective
each type of fitness function using DO, GWO, ALO, and PSO. Again, the

9
F.M. Almasoudi et al. Ain Shams Engineering Journal xxx (xxxx) xxx

Fig. 11. Dynamic response of the power system with the proposed NPI controller compared to literature works in [12,16].

Fig. 12. Objective function analysis for the previous three scenarios.

function due to its superior performance (i.e., Kp = 40.5660, Ki = the system fluctuations. In particular, the coordination strategy-based
32.173, and k = 0.231). fixed ratio has an undershoot of around 0.05 Hz at the beginning of
the scenario, which does not exist in the case of the other adaptive
control strategies. At the load step instant at 50 s, the overshoot in the
4.2. Studied scenarios of the proposed NPI frequency deviation is 0.0008 Hz, 0.03 Hz, and 0.045 Hz with the co­
ordination strategy of fixed ratio, adaptive-based-PI, and adaptive-
Scenario A: Load step change and uniform RESs profiles based-NPI, respectively. Fig. 9e shows the coordination gain Kn with
In this case, a load step change presented in Fig. 4a is considered with the three coordination strategies. The value of Kn changes when the load
uniform solar and wind generation, as shown in Fig. 9b-c. The frequency increases at instant 50 s and the PV power increases to 0.05 pu at instant
deviation of the employed controllers’ techniques, which are fixed ratio, 60 s. The proposed NPI results in a moderate value for the gain Kn
adaptive-based-PI, and proposed adaptive-based-NPI, are depicted in compared to the other coordination strategies, and this Kn value pro­
Fig. 9d. When compared to the other coordinating solutions, it is vides the minimum deviation in the HPS frequency. The FC-ESS power
abundantly obvious that the NPI controller that was presented is capable depends on the RESs power and the gain Kn as shown in Fig. 9f. The
of achieving effective damping for frequency fluctuation. The main profile of the change in the other DG units in the system in this scenario
reason is that the proposed NPI has a nonlinear term that can dampen

10
F.M. Almasoudi et al. Ain Shams Engineering Journal xxx (xxxx) xxx

are shown in Fig. 9g-i for the diesel, EV, and BESS, respectively. These mitigate the mismatch in the HPS parameters due to different aspects.
DGs are controlled directly by the PI-based mMSA, which serves as the Moreover, the results revealed that the proposed strategy could provide
main LFC of the HPS for the three coordination strategies. uniform low power fluctuations from different DG sources, despite
Scenario B: Series load step change and random RESs profiles there-switched power and load variations, where it reduces the fre­
The proposed NPI controller is validated by taking the speculative quency deviation by roughly 95% and 90% compared to the coordina­
variations of the multistep variation of the load demand, as shown in tion strategy based on the fixed ratio and adaptive based-PI,
Fig. 10a. Unlike the previous scenario, here, a random profile in wind respectively. Finally, the proposed strategy could be applied to other
and photovoltaics is used to test the proposed controller against the systems, but the NPI gains must be optimized to handle the system’s
uncertainties of RESs, as shown in Fig. 10b-c. The dynamic response of characteristics.
the frequency deviation is shown in Fig. 10d, considering the three co­
ordination strategies. It is seen that fewer oscillations are observed with CRediT authorship contribution statement
the proposed NPI controller compared to the other controllers during the
simulation time. Based on this, the proposed controller enhances the Fahad M. Almasoudi: Conceptualization, Formal Analysis, Funding
HPS’s robustness and performance under various operating conditions. Acquisition, Methodology, Project administration, Resources, Writing -
Also, the value of the coordination gain Kn during this scenario is shown original draft. Abualkasim Bakeer: Conceptualization, Data curation,
in Fig. 10e. At instant 30 s where the load emphasizes a high step change Formal Analysis, Investigation, Methodology, Software, Validation,
from 0 pu to 0.2 pu, Kn equals zero as the HPS requires the entire RESs Visualization, Writing - original draft. Gaber Magdy: Conceptualiza­
power to substitute for the high power demand of the load. On the other tion, Data curation, Formal Analysis, Investigation, Methodology, Soft­
hand, the power of the FC-ESS reaches zero when Kn = 1 with the pro­ ware, Validation, Visualization, Writing - original draft. Khaled Saleem
posed NPI controller. The power of the other DG units in the HPS is S. Alatawi: Conceptualization, Formal Analysis. Funding Acquisition,
shown in Fig. 10g-i. Resources. Visualization, Writing - original draft. Gaber Shabib: Su­
Scenario C: Parameter mismatches of the HPS pervision, Writing - review and editing. Abderrahim Lakhouit: Fund­
The efficacy of the proposed NPI controller is authenticated by ing Acquisition, Writing-review and editing. Sultan E. Alomrani:
considering fluctuations in the system parameters to study the robust­ Funding Acquisition, Writing - review and editing.
ness of the controller with dynamic changes in the parameters, as shown
in Fig. 11. In Fig. 11d, the frequency deviations are depicted for 50% Funding
uncertainty in the HPS inertia and damping coefficients (i.e., M = 0.2
and D = 0.15). It can be noted that the proposed NPI controller has the This research was funded by the University of Tabuk, grant number
least oscillations in frequency deviation compared to the fixed ratio and S-1443–0207.
adaptive-based PI coordination strategies. It is shown that the proposed https://www.ut.edu.sa/en/Deanship/scientific-research/Pages/def
DO-optimized NPI controller can regulate frequency adequately in HPS, ault.aspx.
even with significant changes to the power system parameters.

4.3. Cumulative objective function evaluation Declaration of Competing Interest

The ISE fitness function is calculated for the three scenarios in which The authors declare that they have no known competing financial
the coordination factor between the RESs and FC-ESS is a fixed ratio, interests or personal relationships that could have appeared to influence
adaptive using the PI controller, or adaptive using the proposed NPI the work reported in this paper.
controller, as shown in Fig. 12. It is obvious that the proposed NPI
controller provides almost negligible value for the cumulative objective Acknowledgments
function compared to other coordination strategies during Scenario A.
Also, the figure demonstrates that the high uncertainty in the RESs The authors extend their appreciation to the Deanship of Scientific
profile, which is the most typical scenario for renewable energy, in­ Research at the University of Tabuk for funding this work through grant
creases the deviation in the system frequency from its nominal value at no. S-1443-0207.
50 Hz (see Scenario B). It can be noted that the proposed NPI-based DO
reduces the frequency deviation by roughly 95% and 90% compared to References
the coordination strategy based on the fixed ratio and adaptive based-PI,
respectively. This proves the superiority of the proposed NPI technique [1] Renewables 2021 Global Status Report (GSR), https://www.ren21.net/reports/gl
obal-status-report/.
in enhancing HPS performance and enabling a high level of RESs [2] Bakeer A, Magdy G, Chub A, Bevrani H. A sophisticated modeling approach for
penetration. photovoltaic systems in load frequency control. Int J Electr Power & Energy Syst.
Elsevier BV 2022;134:107330. https://doi.org/10.1016/j.ijepes.2021.107330.
[3] Qin B, Wang M, Zhang G, Zhang Z. Impact of renewable energy penetration rate on
5. Conclusion power system frequency stability. Energy Reports. Elsevier BV 2022;8:997–1003.
https://doi.org/10.1016/j.egyr.2022.05.261.
This paper proposes a control strategy that employs an NPI controller [4] Abubakar M, Che Y, Ivascu L, Almasoudi FM, Jamil I. Performance Analysis of
Energy Production of Large-Scale Solar Plants Based on Artificial Intelligence
to effectively coordinate energy storage systems (ESS) that are based on (Machine Learning) Technique. Processes 2022;10(9):1843.
fuel cells (FC) and renewable energy sources (RES) to improve the fre­ [5] Sarfraz M, Naseem S, Mohsin M, Bhutta MS. Recent analytical tools to mitigate
quency dynamic performance of hybrid power systems (HPSs) that are carbon-based pollution: New insights by using wavelet coherence for a sustainable
environment. Environ Res 2022;212:113074.
equipped with large-scale RESs. This highly advanced strategy aims to [6] Sarfraz M, Iqbal K, Wang Y, Bhutta MS. Role of agricultural resource sector in
optimize the functionality of the ESS to achieve better performance and environmental emissions and its explicit relationship with sustainable
efficiency in HPSs. The proposed coordination strategy considers the development: Evidence from agri-food system in China. Resour Policy 2023;80:
103191.
system’s nonlinearity by applying the NPI to define the adaptive power-
[7] Bhutta MS, Sarfraz M, Ivascu L, Li H, Rasool G, ul Abidin Jaffri Z, Farooq U, Ali
sharing factor. The NPI parameters are optimally defined using the DO Shaikh J, Nazir MS. Voltage Stability Index Using New Single-Port Equivalent
algorithm. With intensive comparison among the proposed NPI-based Based on Component Peculiarity and Sensitivity Persistence. Processes 2021;9(10):
DO and other coordination strategies based on a fixed ratio and 1849.
[8] Chen W, Liu B, Nazir MS, Abdalla AN, Mohamed MA, Ding Z, et al. An energy
adaptive-based-PI controller, the proposed one demonstrates high effi­ storage assessment: using frequency modulation approach to capture optimal
cacy in all the tested cases. Furthermore, the proposed NPI was able to coordination. Sustainability 2022;14(14):8510.

11
F.M. Almasoudi et al. Ain Shams Engineering Journal xxx (xxxx) xxx

[9] Nazir MS, Abdalla AN, Zhao H, Chu Z, Nazir HMJ, Bhutta MS, et al. Optimized [30] Su YX, Sun D, Duan BY. Design of an enhanced nonlinear PID controller.
economic operation of energy storage integration using improved gravitational Mechatronics. Elsevier BV 2005;15(8):1005–24. https://doi.org/10.1016/j.
search algorithm and dual stage optimization. J Storage Mater 2022;50:104591. mechatronics.2005.03.003.
[10] Magdy G, Bakeer A, Alhasheem M. Superconducting energy storage technology- [31] Shamseldin Mohamed A, Sallam M, Bassiuny AM, Abdel Ghany AM. Real-time
based synthetic inertia system control to enhance frequency dynamic performance implementation of an enhanced nonlinear PID controller based on harmony search
in microgrids with high renewable penetration. Protection Control Modern Power for one-stage servomechanism system. J Mech Eng Sci. Universiti Malaysia Pahang
Syst. Springer Science and Business Media LLC 2021;6(1). https://doi.org/ Publishing 2018;12(4):4161–79. https://doi.org/10.15282/
10.1186/s41601-021-00212-z. jmes.12.4.2018.13.0359.
[11] Pan I, Das S. Fractional Order AGC for Distributed Energy Resources Using Robust [32] El-Gawad AA, Eldeen AN, Bahgat ME, Abdel Ghany AM. Enhanced Non Linear PID
Optimization. IEEE Trans Smart Grid. Institute of Electrical and Electronics Load Frequency Controller Design for a Hydro-Thermal Power System. Proc. IEEE
Engineers (IEEE) 2016;7(5):2175–86. https://doi.org/10.1109/tsg.2015.2459766. International Conference on Electrical Engineering (ICEENG); 2020. doi: 10.1109/
[12] Salama HS, Magdy G, Bakeer A, Vokony I. Adaptive coordination control strategy iceeng45378.2020.9171750.
of renewable energy sources, hydrogen production unit, and fuel cell for frequency [33] Armstrong B, Wade BA. Nonlinear PID Control with Partial State Knowledge:
regulation of a hybrid distributed power system. Protect Control Modern Power Damping without Derivatives. Int J Robot Res 2000;19(8):715–31.
Syst. Springer Science and Business Media LLC 2022;7(1). https://doi.org/ [34] Zhao S, Zhang T, Ma S, Chen M. Dandelion Optimizer: A nature-inspired
10.1186/s41601-022-00258-7. metaheuristic algorithm for engineering applications. Eng Appl Artificial Intell.
[13] Al-Bonsrulah HAZ, et al. Design and Simulation Studies of Hybrid Power Systems Elsevier BV 2022;114:105075. https://doi.org/10.1016/j.engappai.2022.105075.
Based on Photovoltaic, Wind, Electrolyzer, and PEM Fuel Cells. Energies. MDPI AG [35] Casseau V, De Croon G, Izzo D, Pandolfi C. Morphologic and Aerodynamic
2021;14(9):2643. https://doi.org/10.3390/en14092643. Considerations Regarding the Plumed Seeds of Tragopogon pratensis and Their
[14] Bornapour M, Hooshmand R-A, Khodabakhshian A, Parastegari M. Optimal Implications for Seed Dispersal. PLOS One. Public Library of Science (PLoS) 2015;
stochastic coordinated scheduling of proton exchange membrane fuel cell- 10(5):e0125040. https://doi.org/10.1371/journal.pone.0125040.
combined heat and power, wind and photovoltaic units in micro grids considering
hydrogen storage. Appl Energy. Elsevier BV 2017;202:308–22. https://doi.org/
10.1016/j.apenergy.2017.05.133.
[15] Hu J, Shan Y, Xu Y, Guerrero JM. A coordinated control of hybrid ac/dc microgrids Fahad M. Almasoudi received the M.S. and Ph.D. degrees in
electrical engineering from the University of Denver, USA, in
with PV-wind-battery under variable generation and load conditions. Int J Electr
Power Energy Syst. Elsevier BV 2019;104:583–92. https://doi.org/10.1016/j. 2013 and 2018, respectively. He is currently an Assistant Pro­
ijepes.2018.07.037. fessor with the Department of Electrical Engineering, Univer­
[16] Khamari D, Sahu RK, Panda S. A Modified Moth Swarm Algorithm-Based Hybrid sity of Tabuk, Tabuk, Saudi Arabia. His current research
Fuzzy PD–PI Controller for Frequency Regulation of Distributed Power Generation interests include renewable energy, power system, smart grid,
modeling and control of power converters.
System with Electric Vehicle. J Control, Autom Electr Syst. Springer Science and
Business Media LLC 2020;31(3):675–92. https://doi.org/10.1007/s40313-020-
00565-0.
[17] Pan I, Das S. Fractional order fuzzy control of hybrid power system with renewable
generation using chaotic PSO. ISA Trans. Elsevier BV 2016;62:19–29. https://doi.
org/10.1016/j.isatra.2015.03.003.
[18] Kumar Khadanga R, Kumar A, Panda S. Frequency control in hybrid distributed
power systems via type-2 fuzzy PID controller. IET Renewable Power Generat.
Institution of Engineering and Technology (IET) 2021;15(8):1706–23. https://doi.
org/10.1049/rpg2.12140.
[19] El-Fergany AA, El-Hameed MA. Efficient frequency controllers for autonomous Abualkasim Bakeer (S’14) was born in Qena, Egypt in 1990.
two-area hybrid microgrid system using social-spider optimiser. IET Generat, He received the B.Sc. and M.Sc. (Hons.) degrees in Electrical
Transmission & Distribut. Institution of Engineering and Technology (IET) Feb. Engineering from Aswan University, Egypt, in 2012 and 2017,
2017;11(3):637–48. https://doi.org/10.1049/iet-gtd.2016.0455. respectively. In 2014, he joined the Electrical Engineering
[20] Magdy G, Mohamed EA, Shabib G, Elbaset AA, Mitani Y. SMES based a new PID Department, Faculty of Engineering, Aswan University, Aswan,
controller for frequency stability of a real hybrid power system considering high Egypt, first as a demonstrator and then as an assistant lecturer
wind power penetration. IET Renewable Power Generat. Institution of Engineering in 2017. Since September 2019, he has been working towards
and Technology (IET) 2018;12(11):1304–13. https://doi.org/10.1049/iet- his Ph.D. with the Department of Electrical Power Engineering
rpg.2018.5096. and Mechatronics, Tallinn University of Technology, Estonia.
[21] Sobhy MA, Abdelaziz AY, Hasanien HM, Ezzat M. Marine predators algorithm for He is the author/co-author of more than 20 scientific papers. In
load frequency control of modern interconnected power systems including addition, he serves as a reviewer in the IEEE TRANSACTION ON
renewable energy sources and energy storage units. Ain Shams Eng J. Elsevier BV INDUSTRIAL ELECTRONICS, IEEE TRANSACTIONS ON IN­
2021;12(4):3843–57. https://doi.org/10.1016/j.asej.2021.04.031. DUSTRIAL INFORMATICS, and IEEE Journal of Emerging and
[22] Nour M, Magdy G, Chaves-Avila JP, Sanchez-Miralles A, Petlenkov E. Automatic Selected Topics in Industrial Electronics. His main research topics focus on DC-DC con­
Generation Control of a Future Multisource Power System Considering High verters, fault diagnosis and fault tolerance, AC drives, impedance-source power con­
Renewables Penetration and Electric Vehicles: Egyptian Power System in 2035. verters, and model predictive control. Mr. A. Bakeer has been a member of the Power
IEEE Access. Institute of Electrical and Electronics Engineers (IEEE) 2022;10: Electronics Society (PELS) and Industrial Electronics Society (IES) since 2020.
51662–81. https://doi.org/10.1109/access.2022.3174080.
[23] Khooban MH, Gheisarnejad M. A Novel Deep Reinforcement Learning Controller
Based Type-II Fuzzy System: Frequency Regulation in Microgrids. IEEE Trans
Gaber Magdy received B.Sc. and M.Sc. (Hons.) degrees in
Emerg Topics Comput Intell. Institute of Electrical and Electronics Engineers (IEEE)
electrical engineering from Aswan University, Egypt, in 2011
2021;5(4):689–99. https://doi.org/10.1109/tetci.2020.2964886.
and 2014, respectively, and the jointly supervised Ph.D. degree
[24] Fawzy A, Bakeer A, Magdy G, Atawi IE, Roshdy M. Adaptive Virtual Inertia-
in electrical engineering from Minia University, Egypt (Main
Damping System Based on Model Predictive Control for Low-Inertia Microgrids.
university), and Kyushu Institute of Technology, Japan (Host
IEEE Access. Institute of Electrical and Electronics Engineers (IEEE) 2021;9:
university), in 2019. He joined the Electrical Engineering
109718–31. https://doi.org/10.1109/access.2021.3101887.
Department, Faculty of Energy Engineering, Aswan University,
[25] Bevrani H, Feizi MR, Ataee S. Robust Frequency Control in an Islanded Microgrid:
Aswan, Egypt, first as a Demonstrator (Assistant Researcher) in
H∞ and μ-Synthesis Approaches. IEEE Trans Smart Grid. Institute of Electrical and
December 2011, then became an Assistant Lecturer in
Electronics Engineers (IEEE) 2015:1. https://doi.org/10.1109/tsg.2015.2446984.
November 2014. From 2017 to 2019, he was a researcher in the
[26] Singh B, Bishnoi SK, Sharma M, Singh P, Dhundhara S. An application of nature
Department of Electrical and Electronic Engineering at Kyushu
inspried algorithm based dual-stage frequency control strategy for multi micro-grid
Institute of Technology, Japan. He is currently an assistant
system. Ain Shams Eng J. Elsevier BV 2023;14(9):102125. https://doi.org/
professor at the Department of Electrical Engineering, Faculty
10.1016/j.asej.2023.102125.
of Energy Engineering, Aswan University, Aswan, Egypt. His research interests include
[27] Mohamed TH, Alamin MAM, Hassan AM. A novel adaptive load frequency control
power system stability, dynamics and control, digital control, smart/micro-grid control,
in single and interconnected power systems. Ain Shams Eng J. Elsevier BV 2021;12
renewable energies, and energy storage systems.
(2):1763–73. https://doi.org/10.1016/j.asej.2020.08.024.
[28] Ahmed EM, et al. Improving load frequency control performance in interconnected
power systems with a new optimal high degree of freedom cascaded FOTPID-TIDF
controller. Ain Shams Eng J. Elsevier BV 2023;14(10):102207. https://doi.org/
10.1016/j.asej.2023.102207.
[29] Magdy G, et al. Tustin’s technique based digital decentralized load frequency
control in a realistic multi power system considering wind farms and
communications delays. Ain Shams Eng J. Elsevier BV 2019;10(2):327–41. https://
doi.org/10.1016/j.asej.2019.01.004.

12
F.M. Almasoudi et al. Ain Shams Engineering Journal xxx (xxxx) xxx

Khaled Saleem S. Alatawi received the B.S. in electrical en­ Abderrahim Lakhouit received the PhD in Environmental
gineering from King Fahd University of Petroleum and Min­ Engineering. This degree was obtained from University of
erals, Dhahran, Saudi Arabia, in 2008, the M.S. degrees in Sherbrooke, Quebec, Canada. Furthermore, he obtained two
electrical engineering from University of Denver, Denver, CO in master’s degrees in Environmental Engineering and in Renew­
2013, and the Ph.D. degree in electrical engineering from able Energy and Energy Efficiency. One master (5 years of
University of Denver, Denver, CO, in 2019. Now, he is an As­ scholarity) was obtained from University of Liege- Belgium and
sistant Professor with the Electrical Engineering Department, the second from ETS - Montreal, Quebec, Canada. Besides, he
University of Tabuk, Tabuk, Saudi Arabia. has a higher diploma in Marine Science from University of
Liege- Belgium. He is currently an Associate Professor and Ac­
ademic Accreditation Coordinator at University of Tabuk,
Kingdom of Saudi Arabia since 2015. He has more than 40
published papers and three books. he acts as reviewer for many
international journals.

Gaber Shabib, received his B.Sc. degree in electrical engi­ Sultan E. Alomrani received the B.E degree in electrical en­
neering from Al Azhar University. In October 1982 he joins the gineering from King Fahd University of Petroleum & Minerals,
electrical engineering, King Fahad University of Petroleum and KSA, in 2007.He is now preparing a master’s degree in
Minerals, Dhahran Saudi Arabia as a research assistant. In renewable energy engineering at Tabuk University, expected to
December 1985 he received his M.Sc. degree in electrical en­ obtain his master’s degree in June 2023. In addition, he works
gineering from King Fahad University of Petroleum and Min­ as a distribution protection engineering expert at the Saudi
erals. In November 1987 he joins the Qassim Royal Institute, Electricity Company. His research interests include renewable
Qassim, Saudi Arabia as a lecturer. He received his Ph.D. degree energy, electrical sustainability, protection of electrical sys­
from Menoufia University, Egypt, in 2001. He joined Aswan tems, smart grid and power system.
High Institute of Energy, South Valley University, Aswan, Egypt
in 1999. He joined Digital Control Laboratory, Tsukuba Uni­
versity, Japan as a visiting Professor 2006–2007. His research
interests are power system stability, control, Self-tuning con­
trol, Fuzzy logic techniques, Digital control techniques all as applied to power systems.

13