Article 1

Interrelated Solar and Thermal plant Autonomous Generation 2

Control utilizing Metaheuristic Optimization 3

Sanjiv Kumar Jain1,*, Sandeep Bhongade2, Shweta Agrawal3, Abolfazl Mehbodniya4,*, Bhisham Sharma5, Subrata 4
Chowdhury6, Julian L. Webber 4 5

1 Electrical Engineering Department, Medi-Caps University; Indore, India, sanjivkumar.jain@medicaps.ac.in 6

2 Electrical Engineering Department, Shri G. S. Institute of Technology and Science, Indore, India, bhon- 7
gadesandeep@gmail.com 8
3 Institute of Advance Computing, Sage University, Indore, India, Shweta.cse@sageuniversity.in 9
4 Department of Electronics and Communication Engineering, Kuwait College of Science and Technology 10
(KCST), Doha, Kuwait; a.niya@kcst.edu.kw, j.webber@kcst.edu.kw; 11
5 Chitkara University Institute of Engineering and Technology, Chitkara University, Rajpura 140401, Punjab, 12
India; bhisham.pec@gmail.com 13
6 SVCET Engineering College & Technology, Chittoor, Andhra Pradesh 517127, India; subratachow- 14
dhury@svcet.in 15
* Correspondencesanjivkumar.jain@medicaps.ac.in (S.K.J.); a.niya@kcst.edu.kw (A.M.); 16

Abstract: Autonomous control of renewable energy source based two-area thermal generation 17
system has been examined in this study. Using the Solar PV system in one of the control areas cre- 18
ates it a complicated nonlinear system analysis. Step load variation has been considered and the 19
changes in the frequencies and tie line power flow are controlled for zero deviations. A novel Grey 20
Wolf Optimizer, which is a metaheuristic algorithm motivated by the grey wolves is utilized for 21
tuning of the controller gains. The Proportional, Integral, and Derivative gains values are opti- 22
mized for the upgraded two-area Solar Thermal Plant (STP). As, the load connected to the system 23
varies continuously with time, so random load variation is also applied to observe the robustness 24
of the proposed optimization method. Sensitivity analyses have also been adopted with the devia- 25
tion in the time constants of different systems. Inertia constant variations of both the areas are 26
considered from -25% to +25%, with or without STP. The suggested algorithm shows a good dy- 27
namic performance as shown from the simulation results in terms of settling time, over shoot and 28
undershoot. The power in tie line achieves zero deviation quite rapidly in solar based case as 29
Citation:To be added by editorial compared to the without STP. 30
staff during production.

Academic Editor: Firstname Last-

Keywords: Frequency Control; Grey Wolf Optimization; Solar Plant; thermal plant; Control Area 31
name Error; Gain Optimization 32
33
Received: date
Accepted: date
Published: date
1. Introduction 34
Publisher’s Note: MDPI stays neu-
Automatic Generation Control (AGC) is the acute requirement of power plants for 35
tral with regard to jurisdictional
smooth working in the state of load deviations. The control demonstrates that generation 36
claims in published maps and insti-
in power system is automatically regulated agreeing to the changes in loading require- 37
tutional affiliations.
ments. The main theme of AGC study lies in system frequency regulation. The speed falls 38
with increment in load demand and this result into frequency fluctuations [1]. The nature 39
of interrelated solar with the already existing thermal plant is to include the renewable 40
Copyright:© 2022by the authors.
energy sources in the system. Also, the existing grids remain connected with more and 41
Submitted for possible open access
more renewable energy sources. Hence, we have considered the inference of solar PV 42
publication under the terms and
system in one of the control areas. This is the main importance of this work. The necessity 43
conditions of the Creative Commons
Attribution (CC BY) license
of autonomous generation control is one of the prime requirements of the interconnected 44

(https://creativecommons.org/license
or single area system. As, the continuous load variations is faced by the generation facil- 45
s/by/4.0/). ities. They are equipped with the systems (Governors with controllers) to respond like, 46

Energies2022, 15, x. https://doi.org/10.3390/xxxxx www.mdpi.com/journal/energies

Energies 2022, 15, x FOR PEER REVIEW 2 of 22

the primary and secondary controls. In which, generation is altered as per the require- 47
ments through governor controls. Also, it is required to achieve zero steady state devia- 48
tions in frequencies regarding the control areas. The tie power flow (tie-lines) should also 49
have zero steady state error. 50
Now-a-days, the recent power systems are interrelated. So, disruption in one of the 51
area creates disturbance in another areas. Also, the control areas are interconnected 52
through tie lines. Therefore, the objective of AGC is to control both the frequency plus the 53
tie (line) power, at scheduled values. Automatic Generation Control has gained more 54
importance and attention of the researchers after advancement of interconnected system. 55
The main challenge is the frequency instability management. Since, it has a direct con- 56
nection with speed. If the system frequency changes, the speed of a plant changes and 57
hence, the output of the power plant is affected. So, the challenge is to nullify the fre- 58
quency deviations as quickly as possible, after change in the system loading conditions. 59
For fast and efficient controls, the gain of secondary controller must be properly opti- 60
mized for achieving zero valued steady state errors in frequency and tie line power flow. 61
Researchers have adopted several optimization algorithms for gain tuning of secondary 62
controller. Several frequency regulations optimization and control methods have already 63
been established over the decades. 64
The authors have identified deregulation situations, load frequency control opera- 65
tional problems, and their technical solutions regarding standard algorithms required for 66
the deployment of this critical condition in [2]. Parameterized AGC schemes are ex- 67
plored, including non-linear and linear power model structure, classical and optimal 68
control, along with centralized controlled. Wind turbines have also been considered for 69
AGC schemes that utilize intelligent control in [3]. 70
The Differential Evolution (DE) approach based parallel 2-DOF regulator LFC issue 71
is offered. In the same, authors have considered different conditional parameters like 72
generation rate constraints, dead band of governor with delay time in the system mod- 73
eling [4]. Authors presented numerous classical controllers, such as the integral control- 74
ler, for automated generation controls in a hydro-thermal system. Also, sensitivity 75
method is used for estimating the appreciated controller with the optimum parameters 76
and the good robustness, considering large deviations in the system loading [5].To 77
astound the frequency fluctuations related problem, a simple PID control procedure 78
which counterbalances this variability in the system frequency is utilized in [6]. Con- 79
troller parameters have been identified varying along wide spectrum of load varia- 80
tions using an imperialist competitive algorithm for obtaining the optimum response of 81
frequency in the same by the authors. Dye-sensitized solar cells Z-series connected 82
modules are tested in a greenhouse environment to combine the devices' high conversion 83
efficiency, robustness, and transparency is proposed by the authors [7]. Panel statistics 84
are also given through outdoor, along with greenhouse surroundings, in both sunny as 85
well as cloudy conditions are considered, for identifying the trends in efficiency perfor- 86
mance, in the same. An artificial intelligence-based algorithm termed as Hybrid Neu- 87
ro-Fuzzy (HNF) method has been developed by the researchers. The proposed regulator 88
is having the advantage of being able to handle non-linearities. It is also being faster 89
compared to the conventional controllers [8]. Setting of control attributes at optimal 90
values of the power systems centered on PSO aimed at the multi-machines system has 91
been suggested in [9].In the same, author verified that the proposed approach is func- 92
tioning properly in dampening local and inter-area oscillations large variations in load- 93
ing situations and system structures. 94
Researchers are exploring merging of renewable energy technologies for electrical 95
power generation and controls. Authors have shown the implementation of an ANFIS 96
strategy that employs artificial neural networks (ANN) to autonomous generation con- 97
trol of the three imbalanced hydrothermal systems [10]. An independent hybrid gener- 98
ating system operation with solar thermal power, battery energy storage, diesel genera- 99
tors, solar PV, ultra-capacitors, fuel cells, wind turbine-based generators, flywheel, and 100
Energies 2022, 15, x FOR PEER REVIEW 3 of 22

aqua electrolyzer is proposed [11]. Metaheuristic optimization algorithms have become 101
quite prominent in many different spheres of engineering. A combined cycle's automatic 102
generation control Gas turbine generation plant with traditional controllers utiliz- 103
ing Firefly Algorithm parameter tuning is proposed by the authors [12]. For two area 104
thermal system with wide load fluctuations, a Teaching and Learning Based Optimiza- 105
tion based algorithm with 2-Degree of Freedom of PID controller is proposed [13]. For 106
Automatic Generation Control in a deregulated environment, a hybrid-based Teaching 107
Learning Oriented Optimization and Pattern Searching methodology is proposed by the 108
authors, and the findings are compared with published results utilizing Genetic Algo- 109
rithms and Differential Evolution. [14]. 110
Renewable energy with better result attainment for performance of AGC is hosted in 111
the work [15]. One of those algorithms is the Grey Wolf Optimizer (GWO), which is in- 112
spired by the grey wolves [16].In this, the optimizer simulates the key leadership con- 113
figuration and the hunting procedure of the grey wolves. The leadership pyramid is im- 114
itation using four types of grey wolves namely, alpha, delta, beta, and omega. The au- 115
thors presented a mathematical prototypical to assess the impression of trifling photo- 116
voltaic power plants on performance factors considering the economics in a bigger power 117
system [17]. The use of Grey Wolf optimization algorithm for the thermal energy systems 118
incorporated in three control areas for the solution AGC problem is proposed by the 119
authors. In the same, a conventional thermal system with a single stage turbine (reheat) 120
and adequate generation constraints (rate) is considered. Algorithm estimates the per- 121
formances of the proportional plus integral, integral, and PID controllers with good ac- 122
curacies [18]. Authors have suggested a quasi-oppositional dragonfly approach for the 123
tuning of PID control attributes considering three-area model. The results are shown with 124
the help of time absolute error [19]. A mathematical standard of trivial (rooftop) photo- 125
voltaic (PV) generating station has been established by the authors [20]. The AGC solu- 126
tions considering the penetration of the different types of electric vehicles in the electric 127
grids along with the power generation is illustrated by the authors [21-24]. In the research 128
work [25], authors suggested an artificial neural based model utilizing the radial basis 129
function for modelling the non-uniform PV system enactments, while cogitating the 130
frequency deviancy as the input for diverse operating conditions [26]. 131
In the past, de-rating techniques have been utilized to qualify frequency backing 132
functions in the PVs; now, for the first time, de-rating techniques are employed to man- 133
age voltages in the full PV LV distribution systems. [27]. For generating the appropriate 134
patterns of charging for the li-ion batteries, battery modelling and multi-objective con- 135
stricted dynamic programming techniques are proposed. In the same ensemble bioge- 136
ography-based optimization approach is employed for the best solutions [28]. The au- 137
thors suggested AI-based strategies to extend life of the battery from both a production 138
and management standpoint. A critical review of cutting-edge AI-based strategies is also 139
provided considering the approached like ANN and ensemble learning for this control 140
problems [29]. The authors developed an easily understandable ML algorithm for battery 141
manufacturing. It is also demonstrated that the technique can be applied to predict var- 142
ious components of battery capacities, as well as quantifying the dynamic impacts as well 143
as interconnections of the coating factors very efficiently [301]. To adjust the class im- 144
balance and accurately categories three important quality indicators of electrodes, an ef- 145
ficient RUBoost centered classifier framework is proposed. Similarly, experimental find- 146
ings demonstrate that the proposed systems can deal with class inequity difficulties and 147
precisely predicting the characteristics of the produced electrode [31]. Lead less or con- 148
taining little lead to overcome the problems posed by hazardous lead yield halides, 149
perovskite materials are being investigated for use in photovoltaic system by the authors 150
[32]. Research has been undertaken to standardize current AI utilizing deep learning 151
operations in this field after the paper first offers a review of AI as well as the big data in 152
combating COVID-19 [33]. The authors have discussed about using the stacking CQDs of 153
Energies 2022, 15, x FOR PEER REVIEW 4 of 22

various sizes; the graded band orientation approach is employed to reduce the charge 154
carrier diffusion in QNR. [34]. 155
Most of the actual-world optimization is inherently nonlinear and multimodal, with 156
a wide range of complicated constraints. Different goals are frequently at direct conflicts. 157
Even for a unified purpose, approach to optimal solutions is always not possible. Me- 158
taheuristic optimization techniques have gained popularity for solving the complex 159
problems that are otherwise challenging to solve using conventional methods. Discov- 160
ering optimal and even sub-optimal results is not an easy job. 161
Metaheuristic methods are now being used to find high quality results to an ev- 162
er-increasing variety of complex real-world issues, such as combinatorial problems. As, 163
these algorithms can handle multi-objective, multi-solution, and non-stationary problems 164
very efficiently. 165
In the proposed PID controller, the system attributes have been calculated utilizing 166
the metaheuristic algorithm for the evaluation of competitive attributes (parameters). 167
Motivation for this paper is to utilize the metaheuristic algorithm for the tuning of the 168
control parameters of multi area load frequency control with and without insertion of 169
renewable energy sources with wide range of load variations. 170
This is the cutting-edge Grey Wolf Optimizer (GWO) for obtaining admirable 171
transient and steady state performances. The issue of continuous deficiency of the energy 172
possessions is also been addressed in the work. The key contributions of the authors in 173
this article are, 174
1. Integration of the thermal power generations with the renewable generations for the 175
automatic generation controls, considering variable and random step load variation 176
in load demand. 177
2. To achieve the zero-frequency error, after the different load variations and main- 178
taining the system frequency constant at the specified values using the proposed 179
PID controls. 180
3. Maintaining the tie (line) flows at the specified levels for the different changes in 181
load conditions in different generation facilities using the adopted novel me- 182
ta-heuristic method. 183
4. Strong dynamic performance regarding the sensitivity analysis as evidenced by 184
simulation results shown by control routine parameters like, undershoot value, 185
overshoot value, and the settling time. 186

2. Overview of Autonomous Generation Control 187

The system's frequency residues are there considering variations in active power, 188
whereas changes in reactive power have less impact and are mostly determined by vari- 189
ations in the bus voltage gradient. While the automated voltage regulator (AVR) controls 190
the amount of bus voltage and reactive power, the control loop of the load frequency 191
regulator controls real power and frequency. 192
193
In contrast to single area systems, frequency management is a practical issue, which 194
is more frequently related to networks of connected regions. The frequency inaccuracy of 195
one location is related to the issue of other linked areas in big systems. Since, all systems 196
are connected to one another through tie lines. The following traits are connected to the 197
multi-area system: 198
• When a single system loses power, a multi-area system allows electricity to flow 199
from one area to another. In this situation, the blackout problems can be mitigated. 200
• The least amount of frequency variation is also gets aggregated. 201
202
The fundamental drawback of a multi-area system, despite some of its advantages, 203
is that any disturbance in one area will also create disruption in the other control area. 204
Every area needs to be in check for things to run normally and effectively. 205
Energies 2022, 15, x FOR PEER REVIEW 5 of 22

2.1. Two control area system (Thermal Energy) 206

Two control regions for thermal generating systems have been taken into consider- 207
ation. As shown in Figure 1, both control regions having been subjected to step load 208
changes, and governor-turbine systems as their basis. One section has a hard or strong 209
internal connection, while the other has a weak tie line outward connection. Frequency 210
deviances for the areas 1 and 2 are w1 and w2 , respectively. 211

212
Figure 1. Block diagram of two area thermal system. 213

214
All pool members must contribute to the frequency control, without hurting their 215
own net interchange as part of tie line centered bias control scheme. All of this implies 216
that the variation in frequency deviancies and the change in the tie power deviations 217
must have zero value under the steady state conditions. 218
ACE is now the linear function of inter area connected lines power fluctuations and 219
frequency instabilities. The connection provided by equations 1 and 2 illustrates the ex- 220
pressions of ACE for the two control zones, 221
222
ACE 1 = P12 + B1 w1 (1) 223
ACE 2 = P21 + B2 w2 (2) 224

3. Modeling of solar based thermal plant 225

The two most viable sources of renewable power now accessible are solar PV and 226
wind generations. The many solar collector types utilized in sun-based thermal power 227
plants include parabolic trough collectors (PTC), sterling engine (dish), and flat plate 228
collectors (FPC).To get the most solar irradiation, the working fluid is carried by the solar 229
collector. Figure 1 depicts two area systems with thermal power plants. These systems 230
have now been upgraded to include Solar Thermal Plants (STPs), as shown in Figure 2. 231
Area 1 includes STP together, while area 2 simply includes the thermal system. For the 232
thermal system, the nominal attribute values were obtained from [1], and for the solar 233
thermal system, they were taken from [13]. 234
235
Energies 2022, 15, x FOR PEER REVIEW 6 of 22

236
237

Figure 2. Block representation for two area thermal systems with Solar PV. 238

The following abbreviations have been used in the above Figures 1 and 2, 239
𝐵1 and𝐵2 are the constants for frequency biases in area 1 & area 2. 240
𝑅1 and𝑅2 are speed directive attributes for governor of area 1 & area 2. 241
𝑆𝐿𝑃1 , 𝑆𝐿𝑃2 are Load deviations (step changes) in area 1 & area 2. 242
𝐴𝐶𝐸1 , 𝐴𝐶𝐸2 are ACEin area 1 plus area 2. 243
∆𝑃𝑡𝑖𝑒 shows the alteration in tie power flow amid area 1 plus area 2. 244
∆𝜔1 , ∆𝜔2 represent changes in frequency/ frequency aberrations in the area 1 & area 2. 245
246
247

248
Figure 3. Model of collector. 249

In Figure 3 arrow Ti is indicating the direction of fluid flow. A fluid of temperature 250
To will be released. I demonstrate how solar radiation heats pipes and stores energy in 251
working fluid. The Acronyms used here are as follows: 252
253
𝑇𝑜 =Collector(outlet fluid) temperature(˚C), 254
2
𝐼=Solar radiation across collection plane(W/m ), 255

𝑇𝑖 = Collector(inlet fluid) temperature(˚C), 256

Energies 2022, 15, x FOR PEER REVIEW 7 of 22

𝑇𝑒 = Collector(environment) temperature(˚C), 257

3
𝑣= flow rate of collector (or pump) (m /s), 258

𝜂0 = Collector optical efficiency, 259

2
𝑉𝐿 = Collector complete heating loss coefficient(𝑊/𝑚 𝐾), 260

𝑐= Heat capacity (specific) of collector fluid(𝐽/(𝑘𝑔𝐾)), 261

𝐶=(𝜌 ∗ 𝑐 ∗ 𝑉), Heat capacity of the fluid of collector(𝐽/𝐾), 262

2
𝐴= Surface area of the collector (m ), 263
3
𝑉= collector heat transfer volume(m ), 264
3
𝜌= Collector (fluid)corpusdensity(𝑘𝑔/m ), 265

266

3.1. Collector model: 267

𝑑𝑇𝑜(𝑡) 𝐴𝜂0 𝑉𝐿 𝐴 𝑣(𝑡)
= 𝐼(𝑡) − (𝑇𝑎 (𝑡) − 𝑇𝑒 (𝑡)) + (𝑇𝑖 (𝑡) − 𝑇𝑜 (𝑡)) (3) 268
𝑑𝑡
𝐶 𝐶 𝑉
Equation 4 is representing time requirement of variables, 269
𝑇𝑎 (𝑡) = (𝑇𝑖 (𝑡) + 𝑇𝑜 (𝑡))/2 (4) 270
271
Furthermore, a simpler solar collector has been created using a constant fluid flow 272
rate. The value of 𝑇𝑜 (𝑡) from equation 3 is substituted in equation 2. Then equation 2 is 273
arranged in a manner that input parameter would be in left side and output parameters 274
would be right side, so the new equation is as shown in equation 5 275
276
277
𝑑𝑇𝑜(𝑡) 𝑉𝐿 𝐴 𝑣 𝐴𝜂0 𝑣 𝑉𝐿 𝐴 𝑉𝐿 𝐴
+( + ) 𝑇𝑜 (𝑡) = 𝐼(𝑡) + ( − ) 𝑇𝑖 (𝑡) + 𝑇𝑒 (𝑡) (5) 278
𝑑𝑡 2𝐶 𝑉 𝐶 𝑉 2𝐶 𝐶
279
280
After applying Laplace transform to equation (5) 281
282
𝑑𝑇𝑜(𝑡) 𝑉𝐿 𝐴 𝑣 𝐴𝜂0 𝑣 𝑉𝐿 𝐴 𝑉𝐿 𝐴
𝐿[ ]+( + ) 𝐿[𝑇𝑜 (𝑡)] = 𝐿[𝐼(𝑡)] + ( − ) 𝐿[𝑇𝑖 (𝑡)] + 𝐿[𝑇𝑒 (𝑡)] (6) 283
𝑑𝑡 2𝐶 𝑉 𝐶 𝑉 2𝐶 𝐶
284
Outcome of Laplace transform is shown in equation 7, 285
286
𝑉𝐿 𝐴 𝑣 𝐴𝜂0 𝑣 𝑉𝐿 𝐴 𝑉𝐿 𝐴
𝑠𝑇0 (𝑠) − 𝑇𝑜 (0) + ( + ) 𝑇𝑜 (𝑠) = 𝐼(𝑠) + ( − ) 𝑇𝑖 (𝑠) + 𝑇𝑒 (𝑠) (7) 287
2𝐶 𝑉 𝐶 𝑉 2𝐶 𝐶
288
Reorganizing equation 7, we get output input relation as shown in equation 8, 289
290
𝜏 𝜏 𝐴𝜂0 𝜏 𝑣 𝑉𝐿 𝐴 𝜏 𝑉𝐿 𝐴
𝑇𝑜 (𝑠) = 𝑇𝑜 (0) + 𝐼(𝑠) + ( − ) 𝑇𝑖 (𝑠) + 𝑇𝑒 (𝑠) (8) 291
𝜏𝑠+1 𝜏𝑠+1 𝐶 𝜏𝑠+1 𝑉 2𝐶 𝜏𝑠+1 𝐶
292
Here variable τ is representing collector time constant. 293
Output is assumed by considering only input as 𝐼(𝑠), 𝑇𝑒 (𝑠) and 𝑇𝑖 (𝑠)and putting 294
other input zero and neglecting initial conditions, then we get 295
296
297
𝑇𝑜(𝑠) 𝜏 𝐴𝜂0
𝑊1 = = (9) 298
𝐼(𝑠) 𝜏𝑠+1 𝐶
299
𝑇𝑜(𝑠) 𝜏 𝑣 𝑉𝐿 𝐴
𝑊2 = = ( − ) (10) 300
𝑇𝑖 (𝑠) 𝜏𝑠+1 𝑉 2𝐶
301
= 𝑇𝑇𝑜(𝑠𝑠) = 𝜏𝑠+1
( )
𝜏 𝑉𝐿 𝐴
𝑊3 (11) 302
𝑒 𝐶
303
The effect of initial condition onTo (s), 304
Energies 2022, 15, x FOR PEER REVIEW 8 of 22

𝑇𝑜(𝑠) 𝜏
𝑊0 = = (12) 305
𝑇0 (0) 𝜏𝑠+1

Equation 13 is showing individual effect, on equation (10), 306

𝑇𝑜 (𝑠) = 𝑊0 𝑇𝑜 (0) + 𝑊1 𝐼(𝑠) + 𝑊2 𝑇𝑖 (𝑠) + 𝑊3 𝑇𝑒 (𝑠) (13) 307

The transfer rate of heat is assumed to be constant in collector’s model. This transfer 308
function is the output Laplace transform 𝑇0 (𝑠) about the input Laplace transform 309
(𝑇𝑖 (𝑠), 𝐼(𝑠) 𝑎𝑛𝑑 𝑇𝑒 (𝑠)). Outlet temperature’s initial conditions are selected to make solar 310
field model more realistic. 311
312
1
𝜏= 𝑉𝐿 𝐴 𝑣
314
( + )
2𝐶 𝑉
(14) 313
315
Where, variable τ is representing time constant of collector and shown as 𝑇𝑠 . 316
The difference between input (inlet) and the surrounding temperatures is so mini- 317
mal. Equation 15 is showing the function for the solar strength w.r.t. irradiance (solar), 318
319
𝑇𝑜 (𝑠) 𝐾𝑠
𝐺(𝑠) = = (15) 320
𝐼(𝑠) 1+ 𝑇𝑠 𝑠
In the above variable𝐾𝑠 is representing the gain of the collector. 321

4. Objective function 322

The purpose of AGC lies in attaining the value of Control Error zero as fast as pos- 323
sible, as defined in equations (1) and (2). The Time-weighted Integration of the absolute 324
value of deviations (ITAE) is utilized for creating the objective function in present study. 325
It is formulated as the fitness function, as displayed below, 326
t

f =
 w − w .tdt
0
1 2 (16) 327

Here, ∆𝜔1 and ∆𝜔2 are representing frequency deviations in two areas, i.e., area1 328
& area 2. 329
The best result was obtained using this method, which reduces the importance of 330
this function. The optimal value is produced via continuous iterative computing. Repeti- 331
tiveness of the solution is also satisfied to ensure optimal solution. 332
333
5. Grey wolf procedure for optimization 334
The grey wolf method belongs to the class of metaheuristic optimization techniques 335
and is solely employed for calculating the gain constant for optimal outcomes. These 336
Metaheuristic schemes have been castoff successfully to tackle several engineering chal- 337
lenges in the real world. Most of these optimization algorithms were created as a result of 338
the GWO approach, which was developed by a pack of wolves to find prey and develop 339
better hunting techniques, and the survival of the fittest hypothesis of evolutionary al- 340
gorithms. The major reason wolf hunting is the finest is because these wolves’ dwell in 341
packs. Applications of GWO for resolving the issue of economical load dispatch are dis- 342
cussed in [26]. 343
The social structure and collective hunting strategy of the pack of grey wolves have 344
been detailed in this paragraph. This algorithm comes under the metaheuristic group 345
and nature-inspired that replicates the clever search method employed by grey wolves 346
for co-ordination among the group in order to effectively catch prey during hunting. One 347
of the most essential aspects of the grey wolf algorithm is having a rigorous hierarchy at 348
social level which helps in persevering mutual support with each other at the time of 349
hunting. Table 1 shows the social structure of wolves with their responsibilities. 350
351
Energies 2022, 15, x FOR PEER REVIEW 9 of 22

Table 1. Hierarchy level of Wolfs in hunting. 352

Hierarchy Rank Roles

1 α Main authority, imposes a choice on the pack

2 β counselor to Alfa, keep the pack disciplined
3 δ Guardian of wolves who have injured
4 Ω Complies with every pack leader wolves
353
354
5.1. Searching agents 355
α, β, δ and ω are representing the search agents. Here Alpha (α) is representing 356
leader who helps in finding optimal solution during hunting, Beta (β) is representing 357
second level optimal result, Delta (δ) is at third level and has less priority as compare to α 358
and β , omega (ω) follows the solution proposed by α, β and δ keep updating its posi- 359
tion towards the prey [13] . Equations 17 and 18 are used to obtain the optimal solution, 360
361
D = C.Y p (k ) − Y (k ) (17) 362

Y (k + 1) = Y p (k ) − A.D (18) 363

364
Where iteration count is k, Y (k ) and Y p (k ) are the wolf and prey position vectors re- 365
spectively. 366
A = 2b.s1 − b (19) 367

C = 2.s 2 (20) 368

369
s1 along with s 2 are considered as vectors of random nature. The value of b varies 370
between the range 2 to 0. 371
C as well as A shows location of the prey. For C , may have any value in range 0 to 372
2. This indicates weight of the prey. Considering C is bigger than 1, optimal result is 373
almost achieved. The importance of C is basically for locally optimal points avoidance. 374
During hunting obstacles may come, and reflected in the random changes in C . Value of 375
C maintains the uncertain functioning of GWO and through this ambiguity, the best 376
outcome is achieved after specific iterations. 377
A monitors for the prey, since first iteration to last iteration. Firstly, one outcome is 378
believed to be alpha, and then solutions vary from this to find a more suitable result. 379
After best outcome is achieved, convergence is known as to be assisted through A . 380
Divergence occurs when wolves deviate from prey to locate a better prey than this 381
(|A|>1). This modernism in A is finished to ensure procedure skillfulness. 382
Convergence occurs while the result point’s movement is in the best feasible value 383
(1>|A|>-1) [13]. 384
The first iteration is dedicated to divergence, while the second half is dedicated to 385
position update towards a point using alpha along with beta, and delta as the best points. 386
Figure 4accomplishesdeparture from the prey to assimilation to the best ones. 387
388
Energies 2022, 15, x FOR PEER REVIEW 10 of 22

389
Figure 4. Convergence action and divergence action [13]. 390

5.3. Positions appprising 391

It is assumed that alpha possesses awareness of the ideal while omega 392
wolves’ positions are updating. However, beta with the deltas has ideas regarding what 393
is the best solution [13]. Using the defined mathematical calculations, Omega travels to 394
its next ideal position approaching prey, 395

D  = C 1 .Y  (k ) − Y (21) 396

D  = C 2 .Y  (k ) − Y (22) 397

D  = C 3 .Y  (k ) − Y (23) 398

Y 1 = Y  − A1. D  (24) 399

 
Y 2 = Y  − A2 . D  (24) 400
 
 Y (k ) + Y (k ) + Y 3 (k ) )
Y (k + 1) = 
1 2

3
(26) 401

The three best options have a notion about their prey, the other wolves compute 402
their separation from them, that are D  , D  and D  respectively. These distances aid in 403
establishing the three best feasible placements for omega. The next best locations are 404
updated by averaging such three possibilities. 405
This position updating technique is performed until the most optimum point is ob- 406
tained. The following analysis is continued until the maximum iteration has met and the 407
location established after several computations. This process is repeated to ensure that 408
the most recently updated location is the ideal place. 409
410

411
Figure 5. Updation process of the positions of the wolves. 412

In above Figure 5, circles of dissimilar colors are the prey and the wolves. 413
𝑎1 , 𝐶1 𝑎𝑛𝑑 𝑎2 along with 𝐶2 , 𝑎3 𝑎𝑛𝑑 𝐶3 are significant elements for the position of the 414
Energies 2022, 15, x FOR PEER REVIEW 11 of 22

searching agents. The technique drives Omega (hunter) in direction of the estimated lo- 415
cation of prey as displayed in the figure. 416

5.4. Illustration for understanding the GWO procedure 417

GWO has a goal function for which a random set of population is generated. In this 418
instance, the aim is to minimize the function. To commence, a random population is cre- 419
ated, as indicated in Table 2, and then each collection of population functions is com- 420
puted. The functions are shuffled in the first iteration to acquire the top three results, this 421
are displayed in Table 3. Entire leftover solutions are referred to as choices. They are 422
modified in each cycle, as seen below (equations 27, 28 and 29), 423
 itr 
b = 2 − 2  (27) 424
 max itr 
This gives, 425

b = 1.3232 (28) 426

D = 2rand (0,1)[ 4.6917,3.2904]Yb (k ) − Y (29) 427

Table 2. Objective function values for random values of Y1 and Y2. 428

No. (n) Y1 Y2 F
I 5.9202 4.0123 6.9806
II 5.8990 5.0009 6.0031
III 7.4190 7.9874 46.0083
IV 3.0018 1.0103 6.0122
V 5.9089 3.3810 3.9909
VI 5.8902 2.9821 2.9982
VII 6.9708 5.8530 23.9081
VIII 5.0072 2.9802 2.0018
IX 4.7900 2.9907 1.9998
X 5.9702 2.9850 3.0021

Table 3. Best Solutions. 429

Results Y1 Y2 F
for
α 5.0013 2.9084 2.0019
β 5.0108 2.9980 3.0201
δ 6.0015 2.9581 3.2039

Y1 = [5.0013,2.9084] − (2b..rand (0,1) − b) D  (30) 430

D  = 2rand (0,1)[5.0108,2.9980]Yb (k ) − Y (31) 431

X 2 = [5.0108,2.9980] − (2b.rand (0,1) − b) D  (32) 432

D  = 2rand (0,1)[6.0015,2.9581]Y (k ) − Y (33) 433

b

Y3 = [6.0015,2.9581] − (2b.rand (0,1) − b) D  (34) 434

  
 Y (k ) + Y2 (k ) + Y3 (k )
Y (1, :) = 1 (35) 435
3
Energies 2022, 15, x FOR PEER REVIEW 12 of 22


Y = [3.0480 ,2.6845 ] (36) 436
The initial value combination's position is updated here. Some is done for 437
the alternative solutions. If the value of omega in the following iteration function is lesser 438
than alpha, now these solutions for omega becomes the subsequent best solution. 439

f omega  f alpha (37) 440

This technique is repeated until the minimal function value is reached. Figure 6 de- 441
picts the flowchart for the suggested Grey Wolf Optimization procedure. 442

443

Figure 6. Flow diagram of the proposed algorithm. 444

6. Result and Discussions 445

Results and the corresponding discussions are given as follows, 446

6.1. Result with STP and without STP 447

In order to compare the effects of solar energy on the system's different characteris- 448
tics in both scenarios with and without STP, solar energy is employed in the power sys- 449
tem. The Table 4 reflects the customized and optimized values for both cases. 450
451
452
453
454
455
Energies 2022, 15, x FOR PEER REVIEW 13 of 22

Table 4. Controller gains without and with STP. 456

Gain
457 of controllers Non-STP STP
k p ,1 0.5601 0.7498 458

459 k p,2 0.7898 0.3311

460
k i ,1 0.4084 0.3202
461
k i,2 0.4802 0.3402
462
k d ,1 0.4983 0.5387
463

464 kd ,2 0.3456 0.7008

465
As shown in Figure 7, the frequency variation area 1 is reaching quickly to zero in 466
the case of a solar thermal power plant. Figure 8displays frequency discrepancy in the 467
area 2 and demonstrated improved results in case of using with PV.Figure 9, displays 468
power flow in the tie line without and with STP. SLP value considered here is 0.2. (Here, 469
TH SOURCE-Thermal source and STP SOURCE – Solar thermal power source) 470
471
0.5 X10-2

∆w1
TH SOURCE
STP SOURCE

-1

-2

0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 X10

Time (s)
472
Figure 7. Changesin frequency in area 1 without and with STP (SLP two percent). 473

474
0 X10-2

TH SOURCE
-0.4 STP SOURCE

∆w2
-0.8

-1.2

-1.6
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 X10
Time (s)
475
Figure 8. Changesinfrequency in area 2 without and with STP (SLP two percent). 476
Energies 2022, 15, x FOR PEER REVIEW 14 of 22

477
x 10-4
5

-0.5 TH SOURCE
ΔPtie STP SOURCE

-1.0

-15

-20
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 X10
Time (s)

478
Figure 9. Power flow in tie line without and with STP (SLP two percent) 479

In the solar thermal instance, fluctuations in the power flow of tie line attain the de- 480
sired null value in the least time. Table 5 has correlation between dynamic performances 481
of both the systems. It can be concluded that solar power helps in improving dynamic 482
performance of system. 483

Table 5. Performance attributes dynamics without and with STP (Undershoot-US, Overshoot-OS, 484
and Time to settle-TS) . 485

Attributes Performance Without STP Along with STP

(dynamics)
TS 21.500312 8.649830
∆w1 OS 0.0107588 0.001642
US 0.0007589 0.001342
TS 20.989801 9.988071
∆w2 OS 0.0106175 0.008796
US 0.0004987 0.000543
TS 22.9012341 9.501261
∆Ptie OS 0.0004598 0.005768
US 0.00019978 0.0003650

6.2. Comparison of results obtained from GWO with Integral (I) controllers 486
The steady state error is reduced by using integral control. Results are showing 487
comparison of integral control with PID control. Grey Wolf Optimization (GWO) is used 488
to optimize attributes of PID controller. Table 6 is showing comparison of settling time, 489
undershoot, and overshoot for various parameter characteristics. It shows that results 490
produced by GWD tuning are the much better as compare to dynamic and I tuning. 491
492
493
494
495
496
497
498
499
Energies 2022, 15, x FOR PEER REVIEW 15 of 22

Table 6. Performance of GWO tuning Vs integral tuning (Undershoot-US, Overshoot-OS, and Time 500
to settle-TS). 501

502
Attributes Dynamics I tuning GWO tuning

TS 24.17865 8.654140
∆w1 OS 0.034786 0.002642
US 0.0076399 0.001242
TS 22.64792 10.075095
∆w2 OS 0.025696 0.008096
US 0.0086329 0.0004932
TS 22.98635 9.48519
∆Ptie OS 0.0065920 0.006068
US 0.0089521 0.0004010
503
Jaya algorithm is used to tune PID for STP for comparisons [35]. Table 7 is showing 504
the comparison between I control, PI control and tuned PID control. 505
506

Table 7. Comparison of I and PI controllers with tuned PID for STP (Undershoot-US, Over- 507
shoot-OS, and Time to settle-TS). 508

Attrib- Dynamic I controller PI controller Tuned PID

utes control
TS 11.92054 14.418729 8.6182208
∆w1 OS 0.013824513 0.0123271884 0.003787487
US 0.035138888 0.021846066 0.021198573
TS 14.658047 13.899727 7.6016538
∆w2 OS 0.0012796318 0.0016410739 0.00003422029
US 0.017074277 0.020000111 0.014809222
TS 24.370994 31.009171 22.898826
∆Ptie OS 0.014122517 0.0040503251 0.00167344
US 0.045736824 0.0087653243 0.012513358
509
The below Figures 10, 11 and 12 give the analogy of dynamic parameters in chart 510
pattern, 511
512
Energies 2022, 15, x FOR PEER REVIEW 16 of 22

Settling time PID with GWO

30
I control
24.17865 22.64792 22.98635
25

20

15
8.65414 10.0750 9.48519
10

0
dw1 dw2 tie line power 513
Figure 10. Settling time variations in I and PID controllers. 514

515

Overshoot PID with GWO

0.04
I control
0.035 0.034786

0.03
0.025696
0.025
0.02
0.015
0.00809
0.01
0.006592
0.005
0.00264 0.006068
0
dw1 dw2 tie line power
516
Figure 11. Overshoot variations in I and PID controllers. 517

518 PID with GWO

519
Undershoot I control
0.01
520
0.009 0.0086329 0.0089521
521
0.008 0.0076399
0.007
0.006
0.005
0.004
0.003
0.002 0.00124
0.001 0.000493 0.00040
0
dw1 dw2 tie line power
Figure 12. Undershoot variations in I and PID controllers. 522

523
Energies 2022, 15, x FOR PEER REVIEW 17 of 22

6.3. Variation in SLP 524

Figure 13 shows the graph of variable SLP with respect to time. It is showing better 525
results as compare to without STP. Incorporation of solar power is giving better results 526
with continuously varying SLP also. A variable step (random) load changes are deliber- 527
ated in area 1 with STP. 528
529
530

SLP
X10-2

18

14

0 0.1 0.2 0.3 0.4 X10

Time (sec)
531
Figure 13. SLPs variations with time scale. 532

Robustness of the offered AGC scheme with STP is displayed in Figures 14 to 19. 533
The results shows that better results are presented by GWO based PID controllers. (Here, 534
TH SOURCE-Thermal source and STP SOURCE – Solar thermal power source) 535
Δω1 x 10-3
-
8

-2

-4 TH SOURCE
STP SOURCE
-6

-8

-10
0 5 10 15 20 25 30 35 40
Time(s)
536
Figure 14. Frequency variations of area 1 without and with STP with variable SLP. 537
Energies 2022, 15, x FOR PEER REVIEW 18 of 22

538
-3
x 10
8

6 TH SOURCE
STP SOURCE
4

2
Δω2
0

-2

-4

-6

-8
0 5 10 15 20 25 30 35 40
Time(s)

539
Figure 15. Frequency variations of area 2 without and with STP with variable SLP. 540

X10-3
5
TH SOURCE
STP SOURCE

0
∆Ptie

-5

-10

-15
0 5 10 15 20 25 30 35 40
Time(s)
541
Figure 16. Power flow fluctuations in tie line without and with STP with variable SLP. 542

Settling time without STPP

45
38.518625 38.210954 38.171784
40
with STPP
35 32.421622
29.995905 30.403231
30
25
20
15
10
5
0
dw1 dw2 tie line power
543
Figure 17. Variations of settling time with step load perturbation. 544

545
Energies 2022, 15, x FOR PEER REVIEW 19 of 22

Undershoot Without STPP

0.5
0.4384 With STPP

0.4

0.3

0.2
0.1302 0.15144
0.09826
0.1 0.0766
0.05167

0
dw1 dw2 tie line power
546
Figure 18. Undershoot variations with step load perturbation. 547

Overshoot
Without STPP
0.6

0.4764 With STPP

0.5

0.4

0.3

0.2 0.16639
0.1301
0.1
0.05477 0.03910.04164
0
dw1 dw2 tie line power
548
Figure 19. Overshoot variations through step load agitation. 549

6.4. Sensitivity investigation 550

Suggested AGC system is evaluated through sensitivity investigations also. Diverse 551
loading scenarios and varied parameters in safe limits are used to analyze system’s sen- 552
sitivity. The system without and with STP is used in wide range of cases due to the better 553
performance. Sensitivity analysis is done by varying the parameters from Figure 2. 554
Fluctuations in the range of -25% to +25%. In all parameter are used and shown in Table 555
8. 556
557
558
559
560
561
562
563
564
565
566
567
Energies 2022, 15, x FOR PEER REVIEW 20 of 22

Table 8. Sensitivity investigation results (X = +25% and Y =-25%). 568

Attribute % Variation ∆w1 ∆w2 ∆Ptie

X 7.4991 6.9988 9.5012
H1
Y 8.0112 9.6014 15.0025
X 8.0109 9.0192 8.9873
Tg1
Y 9.0017 9.9905 10.9797
X 6.7002 8.9782 13.9826
Tt1
Y 10.00534 10.9936 11.9837
X 10.998 11.9935 18.0088
Ts
Y 7.9842 6.0045 5.7681
Table 8 shows the attributes regarding settling time; ITAE and overshoot (peak) are 569
varying in safe limits and are almost equal to nominal values of system attributes. As a 570
result, the given procedure delivers satisfactory, stable, robust control, along with the 571
optimal results of the regulator attributes acquired at standard loading along with set 572
values do not need resetting for extensive variations in loading or the parameters. 573

7. Conclusions 574
The simulated trainings from the proposed strategy in Solar Thermal Plant (STP) 575
are compared with various regulating techniques. The findings using GWO along with 576
PID reflected better transient performance comparing with simple control (I-control). A 577
comparison analysis for the developed framework without plus with STP was conduct- 578
ed, and simulation outcomes demonstrate that the integration of solar PV, as type of re- 579
newable sources, enhanced system results completely. Step load agitation implies vary- 580
ing the percentage adjustments in load to test the resilience of system operation and ob- 581
tain all performance attributes that are the best in line with the wide variations of step 582
load fluctuation. Variable SLP is also investigated to consider the situations more realis- 583
tic. All system parameters Tg1, Tg2, H, Tt1, Tt2 and Ts were adjusted with a 25% augmenta- 584
tion and reduction for sensitivity study of autonomous generation control using GWO. 585
Because a 25% fluctuation is probable, the algorithm's sensitivity is tested for this change. 586
The limitation of the present work is regarding the convergence speed, which may be 587
further improved by using dynamic weights. 588
In the present work, incorporation of the solar PV with thermal plant by the sub- 589
mission of GWO boosted controllers for the two area systems is deliberated. The work 590
may be extended in future with the insertion of EVs with AGC of multi area systems for 591
optimal control parameters tuning. Also, researchers may adopt hybrid optimization 592
algorithms for optimal parameter setting with inclusion of wind sources and storage. 593
594
Acknowledgement: The authors would like to thank Kuwait College of Science & Tech- 595
nology (KCST) for technical and financial support of this research. 596
597
Author Contributions: “Conceptualization, Sandeep Bhongade, Sanjiv Kumar Jain and Abolfazl 598
Mehbodniya.; methodology, Sandeep Bhongade, Shweta Agrawal and Bhisham Sharma; software, 599
Sandeep Bhongade, Sanjiv Kumar Jain and Shweta Agrawal ; validation, Sandeep Bhongade, Julian 600
L. Webber, Abolfazl Mehbodniyaand Sanjiv Kumar Jain; formal analysis, Sandeep Bhongade, 601
Subrata ChowdhuryandAbolfazl Mehbodniya; investigation, Sandeep Bhongade, Sanjiv Kumar 602
Jain, Shweta Agrawal and Subrata Chowdhury; resources, Julian L. Webber, Abolfazl Mehbodni- 603
yaand Sanjiv Kumar Jain; data curation, Sandeep Bhongade, Abolfazl Mehbodniyaand Sanjiv Ku- 604
mar Jain.; writing—original draft preparation, Sandeep Bhongade, Bhisham Sharma, Abolfazl Me- 605
hbodniyaand Sanjiv Kumar Jain; writing—review and editing, Sandeep Bhongade, Bhisham Shar- 606
ma, Abolfazl Mehbodniyaand Subrata Chowdhury; visualization, Sandeep Bhongade, Sanjiv Ku- 607
mar Jain and Subrata Chowdhury; supervision, Sandeep Bhongade, Sanjiv kumar Jainand Abolfazl 608
Mehbodniya; project administration, Sandeep Bhongade, Sanjiv Kumar Jain and Bhisham Sharma; 609
funding acquisition, Abolfazl Mehbodniya and Julian L. Webber. 610
Energies 2022, 15, x FOR PEER REVIEW 21 of 22

All authors have read and agreed to the published version of the manuscript. 611

Funding: This research received no external funding. 612

Institutional Review Board Statement: Not applicable. 613

Informed Consent Statement: Not applicable. 614

Data Availability Statement: Not applicable. 615

Conflicts of Interest: The authors declare no conflict of interest. 616

References 617
1. Elgerd, O.I.; Fosha, C. The Megawatt-Frequency Control Problem: A New Approach via Optimal Control Theory.IEEE Trans. 618
Power Apparatus System1970, PAS-89, pp. 563–577. 619
2. Christie, R.D.; Bose, A. Load Frequency Control Issues in Power System Operation after Deregulation. IEEE Trans. on Power 620
Systems, 1996, 11, pp. 1191-1200. 621
3. Kumar, P.; Kothari, D.P. Recent philosophies of automatic generation control strategies in power systems. IEEE Trans Power 622
Syst, 2005, 20, pp. 346-357. 623
4. Kumar, S.R.; Panda, S.; Rout U.K. DE optimized parallel 2-DOF PID controller for load frequency control of power system with 624
governor dead-band nonlinearity. Electr Power Energy Syst, 2013, 49, pp. 19-33. 625
5. Chandra, S.L.; Nanda, J.; Mishra, S. Performance comparison of Several classical controllers in AGC for multi-area intercon- 626
nected thermal system. Int J Electr Power Energy Syst 2011, 33, pp. 394-401. 627
6. Hamed,S.; Behrooz, V.; Majid, E. A robust PID controller based on imperialist competitive algorithm for load-frequency con- 628
trol of power systems. ISA Trans 2013, 52. pp. 88-95. 629
7. Barichello, J.; Vesce, L.; Mariani, P.; Leonardi, E.; Braglia, R.; Di Carlo, A.; Canini, A.; Reale, A. Stable semi-transparent 630
dye-sensitized solar modules and panels for greenhouse application. Energies 2021, 14, p.6393. 631
8. Panda, G.; Panda, S.; Ardil, C., Hybrid neuro fuzzy approach for automatic generation control of two-area interconnected 632
power system. International Journal of Computational Intelligence, 2009, 5, pp.80-84. 633
9. Abido, M.A. Optimal design of power-system stabilizers using particle swarm optimization. IEEE Trans Energy Conversion 634
2002, 17, pp.406-13. 635
10. Khuntia, S.R.; Panda, S. Simulation study for automatic generation control of a multi-area power system by ANFIS approach. 636
Applied soft computing 2012, 12, pp.333-341. 637
11. Das, D.C.; Roy, A.K.; Sinha, N. GA based frequency controller for solar thermal-diesel-wind hybrid energy generation/energy 638
storage system. Int JElectr Power Energy Syst 2012, 43, pp.262-279. 639
12. Saikia, L.C.; Sahu,S.K. Automatic generation control of a combined cycle Gas Turbine plant with classical controllers using 640
Firefly Algorithm. Electr Power Energy Syst 2013; 53, pp.27-33. 641
13. Sahu, R.K.; Panda, S.; Rout, U.K.; Sahoo, D.K. Teaching learning based optimization algorithm for automatic generation control 642
of power system using 2-DOF PID controller, Int J Elect Power Energy Syst 2016, 77, pp. 287-301. 643
14. Khamari, D.; Sahu, R.K.; Gorripotu, T.S.; Panda, S. Automatic generation control of power system in deregulated environment 644
using hybrid TLBO and pattern search technique. Ain Shams Eng Journal 2020, 11, pp. 553-573. 645
15. Buzás, J.; Kicsiny, R. Transfer functions of solar collectors for dynamical analysis and control design. Renewable Energy 2014; 646
68, pp.146-55. 647
16. Mirjalili, S.; Mirjalili, S.M.; Lewis, A. Grey wolf optimizer. Advances in engineering software 2014, 69, pp.46-61. 648
17. Asano, H.; Yajima, K.; Kaya, Y.; Influence of photovoltaic power generation on required capacity for load frequency control. 649
IEEE Trans Energy Convers 1996;11, pp.188-93. 650
18. Sharma, Y.; Saikia, L.C. Automatic generation control of a multi-area ST-Thermal power system using Grey Wolf Optimizer 651
algorithm based classical controllers. International Journal of Electrical Power & Energy Systems 2015, 73, pp.853-862. 652
19. Vedik, B.; Kumar, R.; Deshmukh, R..; Verma, S.; Shiva, C.K. Renewable energy-based load frequency stabilization of inter- 653
connected power systems using quasi-oppositional dragonfly algorithm. Journal of Control, Automation and Electrical Sys- 654
tems 2021, 32, pp. 227-243. 655
20. Reddy, V.S.; Kaushik, S.C.; Ranjan, K.R.; Tyagi, S.K. State-of-the-art of solar thermal power plants-A review. Renewable and 656
Sustainable Energy Reviews 2013, 27, pp.258-273. 657
21. Babaei, F.; Safari, A. SCA based fractional-order PID controller considering delayed EV aggregators. Journal of Operation and 658
Automation in Power Engineering 2020, 8, pp.75-85. 659
22. Liu, H..; Yang, Y.; Qi, J.; Li, J.;Wei, H.; Li, P. Frequency droop control with scheduled charging of electric vehicles. IET Gener- 660
ation, Transmission & Distribution 2017, 11, pp.649-656. 661
23. Latif, A.; Pramanik, A.; Das, D.C.; Hussain, I.; Ranjan, S. Plug in hybrid vehicle-wind-diesel autonomous hybrid power system: 662
frequency control using FA and CSA optimized controller. International Journal of System Assurance Engineering and Man- 663
agement 2018, 9, pp.1147-1158. 664
24. Pham, T.N.; Trinh, H. Load frequency control of power system with an electric vehicle and diverse transmission links using 665
distributed functional observers. IEEE Trans. Smart Grid 2016, 7, pp. 238-252. 666
Energies 2022, 15, x FOR PEER REVIEW 22 of 22

25. Bakeer, A.; Magdy, G.; Chub, A.; Bevrani, H.; A sophisticated modeling approach for photovoltaic systems in load frequency 667
control. Int J Electr Power Energy Syst 2022, 134, p.107330. 668
26. Pradhan, M.; Roy, P.K.; Pal, T. Grey wolf optimization applied to economic load dispatch problems. Int J Electr Power Energy 669
Syst 2016, 83, pp.325-334. 670
27. Verma P, Katal N, Sharma B, Chowdhury S, Mehbodniya A, Webber JL, Bostani A. Voltage Rise Mitigation in PV Rich LV 671
Distribution Networks Using DC/DC Converter Level Active Power Curtailment Method. Energies. 2022, 15, p. 5901. 672
28. Liu, K.; Zou, C.; Li, K.; Wik, T. Charging pattern optimization for lithium-ion batteries with an electrothermal-aging model. 673
IEEE transactions on industrial informatics 2018, 14, pp.5463-5474. 674
29. Liu, K.; Wei, Z.; Zhang, C; Shang, Y.; Teodorescu, R.; Han, Q.L. Towards long lifetime battery: AI-based manufacturing and 675
management. IEEE/CAA Journal of AutomaticaSinica 2022. 676
30. Liu, K.; Niri, M.F.; Apachitei, G.; Lain, M.; Greenwood, D.; Marco, J. Interpretable machine learning for battery capacities pre- 677
diction and coating parameters analysis. Control Engineering Practice 2022, 124, p.105202. 678
31. Liu, K.; Hu, X.; Meng, J.; Guerrero, J.M.; Teodorescu, R. RUBoost-based ensemble machine learning for electrode quality clas- 679
sification in Li-ion battery manufacturing. IEEE/ASME transactions on mechatronics 2021. 680
32. Pandey, R.; Sharma, S.; Madan, J.; Sharma, R. Numerical simulations of 22% efficient all-perovskite tandem solar cell utilizing 681
lead-free and low lead content halide perovskites. Journal of Micromechanics and Microengineering 2021, 32, p.014004. 682
33. Manoj, A.; Shweta, A. A systematic review on artificial intelligence/deep learning applications and challenges to battle against 683
covid-19 pandemic. Disaster Advances 2021, pp.90-99. 684
34. Pandey, R., Khanna, A., Singh, K., Patel, S.K., Singh, H. and Madan, J., 2020. Device simulations: Toward the design of> 13% 685
efficient PbS colloidal quantum dot solar cell. Solar Energy, 207, pp.893-902. 686
35. Bhongade, S.; Paramar, V.P. Automatic generation control of two-area ST-Thermal power system using jaya algorithm. Int. J. 687
Smart Grid 2018, 2, pp.99-110. 688
689
690
691