An improved search ability of particle swarm optimization algorithm for tracking maximum power point under shading conditions Текст научной статьи по специальности «Электротехника, электронная техника, информационные технологии»

UDC 621.314

https://doi.org/10.20998/2074-272X.2022.2.04

H. Saeed, T. Mehmood, F.A. Khan, M.S. Shah, M.F. Ullah, H. Ali

An improved search ability of particle swarm optimization algorithm for tracking maximum power point under shading conditions

Introduction. Extracting maximum possible power from solar energy is a hot topic of the day as other sources have become costly and lead to pollution. Problem. Dependency on sunlight for power generation makes it unfeasible to extract maximum power. Environmental conditions like shading, partial shading and weak shading are the major aspect due to which the output of photovoltaic systems is greatly affected. Partial shading is the most known issue. Goal. There have been many proposed techniques and algorithms to extract maximum output from solar resources by use of photovoltaic arrays but every technique has had some shortcomings that couldn't serve the complete purpose. Methodology. Nature inspired algorithms have proven to be good to search global maximum in a partially shaded multipeak curve which includes particle swarm optimization, artificial bee colony algorithm, and flower pollination algorithm. Methods. Particle swarm optimization algorithm is best among these in finding global peaks with less oscillation around maximum power point, less complexity, and easy to implement nature. Particle swarm optimization algorithm has the disadvantage of having a long computational time and converging speed, particularly under strong shading conditions. Originality. In this paper, an improved opposition based particle swarm optimization algorithm is proposed to track the global maximum power point of a solar photovoltaic module. Simulation studies have been carried out in MATLAB/Simulink R2018a. Practical value. Simulation studies have proved that opposition based particle swarm optimization algorithm is more efficient, less complex, more robust, and more flexible and has better convergence speed than particle swarm optimization algorithm, perturb and observe algorithm, hill climbing algorithm, and incremental conductance algorithm. References 24, tables 4, figures 12. Key words: conventional particle swarm optimization, maximum power point, opposition based particle swarm optimization algorithm.

Вступ. Отримання максимально можливоi потужностi i3 сонячноi енергп е надзвичайно актуальним Hapa3i, осюльки iншi джерела енергп стали коштовними та призводять до забруднення. Проблема. Залежтсть eid сонячного свтла для вироблення електроенергп унеможливлюе отримання мaксимaльноi потужностi. Умови навколишнього середовища, таю як заттення, часткове заттення i слабке заттення, е основним аспектом, вiд якого сильно залежить потужтсть фотоелектричних систем. Часткове заттення - нaйвiдомiшa проблема. Мета. Було запропоновано багато методiв та aлгоритмiв для отримання мaксимaльноi вiддaчi вiд сонячних ресурЫв за допомогою фотоелектричних батарей, але кожен метод мав деяш недолжи, ят не могли служити досягненню повноi мети. Методологы. Алгоритми, натхнент природою, виявилися хорошими для пошуку глобального максимуму на частково заттенш кривш з багатьма тками, включаючи оптимiзaцiю рою частинок, алгоритм штучноi бджолиноi колони та алгоритм запилення квтв. Методи. Алгоритм оптимiзaцii рою частинок найкраще тдходить для пошуку глобальних тюв з меншими коливаннями навколо точки мaксимaльноi потужностi, меншою складшстю та простотою реaлiзaцii. Алгоритм оптимiзaцii рою частинок мае недолт, що полягае у тривалому чаЫ обчислень та ms^^^i збiжностi, особливо в умовах сильного заттення. Оригшальшсть. У цш стaттi пропонуеться покращений алгоритм оптимiзaцii рою частинок на основi протилежностi для вiдстеження глобaльноi точки мaксимaльноi потужностi сонячного фотоелектричного модуля. Розрaхунковi моделювання проводились у MATLAB/Simulink R2018a. Практична цшн^ть. До^дження за допомогою моделювання довели, що алгоритм оптимiзaцiiрою частинок на основi протилежностi е быьш ефективним, менш складним, надшшшим i гнучюшим i мае кращу швидюсть збiжностi, тж алгоритм оптимiзaцii рою частинок, алгоритм збурення та спостереження, алгоритм сходження на пагорб та алгоритм iнкрементaльноi провiдностi. Бiбл. 24, табл. 4, рис. 12. Ключовi слова: традицшна оптимiзацiя рою частинок, точка максимально!" потужносп, алгоритм onraMi3a^i рою частинок на основi протилежность

Abbreviations

ABC Artificial Bee Colony OBL Opposition Based Learning

FPA Flower Pollination Algorithm P&O Perturb and Observe Algorithm

GMPP Global Maximum Power Point PSC Partial Shading Conditions

IC Incremental Conductance Algorithm PSO Particle Swarm Optimization

IPSO Improved Particle Swarm Optimization PSO-OBL Opposition Based Particle Swarm Optimization Algorithm

MPP Maximum Power Point PV Photovoltaic

MPPT Maximum Power Point Tracking PWM Pulse-Width Modulation

1. Introduction. With increasing demand of renewable source of energy, solar energy is getting more and more attention because of its easy and never ending availability. But operating PV panel at MPPT is a challenging task. So accomplishing this task can make use of PV panels very effective and it will largely reduce the use of non-renewable sources which will have such a huge impact on mankind's lifestyle that it can save a huge amount of money and world resources. However high installation cost and getting lower efficiency are still the challenges that are faced nowadays by researchers [1, 2]. Solar energy plays a vital role in the development of a country. Compared to other renewable sources solar

energy generation shows a significantly more dynamic development. Research studies shows that the installed capacity of solar power plants in 2016 was 315 GW [3-6]. For extracting MPPT, there are face several challenges such as temperature and irradiance affects the P-V characteristics of photovoltaic cell. For delivering maximum power to load several conventional algorithms were to tackle MPPT. For instance P&O MPPT method [7], IC MPPT method [8], modified hill climbing method [9], fictional short-circuit current approach [10], ripple correlation control method [11] and sliding mode controller [12] were used. The specialty of these

© H. Saeed, T. Mehmood, F.A. Khan, M.S. Shah, M.F. Ullah, H. Ali

algorithms is that these give good response for uniform irradiance of solar panel but under PSC these algorithms fails to track the global MPPT.

Goal. This paper present a hybrid technique used to track global max power point very efficiently by using improved particle swarm optimization algorithm. For improving search ability of PSO algorithm, PSO is hybridized with OBL [13], for getting high tracking efficiency, convergence speed and power output. PSO is an outstanding choice for problems of nature of non-linear optimization [14]. To check the robustness of this hybrid techniques, two different PV configurations of are employed.

The PV system consists of PV panels, MPPT controller, DC-DC converter and load (Fig. 1). The output of the PV panels is fed to the converter as well as to the controller. The controller uses its input to regulate the duty cycle of the converter for maximum power point operation.

Table 1

Parameters of Sun Earth Solar Power TPZ250MBZ PV 250 W

Fig. 1. Block diagram of PV system

This paper is organized as follows. In Section 2 there is description of P-V characteristics under PSC. In Section 3 methodology is describe. In Section 4 test system is presented while in Section 5 simulation results are discussed. In Section 6, comparison is discussed and paper is concluded in Section 7.

2. PV cell's modeling with single-diode equivalent circuit.

2.1. Characteristics of solar cell. Equivalent circuit of PV cell is shown in the Fig. 2 and the output in the form of current is given by:

1 pv 1 ph I0

U pv -1 ' RS

exp

q

('Upv

+I • R

S

A • K • T

-1

R

where Ipv is the output current of cell; Iph is the photocurrent; I0 is the reverse saturation current; q is the charge of an electron; Upv is the work voltage of a PV cell; K is the Boltzmann constant; A is the ideality factor; RS and RP are the series and parallel resistances; T is the cell temperature [15].

The different parameters of proposed PV module are shown in Table 1.

Im

pv

m:

Rs

Parameters Variables Values

Maximum power PMPP 250 W

Open circuit voltage V y oc 30 V

Short circuit current ISC 8.3 A

Current of Pmax Impp 8.83 A

Voltage of Pmax VMPP 36.8 V

Isc coefficient of temperature K (0.065±0.015)%/°C

Voc coefficient of temperature Kv -(0.32±0.05)%/°C

2.2. Effect of PSC on PV characteristics. In this paper, two different PV configurations are used to check the robustness of proposed technique. The first configuration is 4S2P, which consists of 8 PV modules. Four modules are organized in one string with power of 249 W. The overall power extracted is 1992 W. Second configuration is 3S1P, which consists of 3 PV modules that are arranged in one string. The overall power extracted by second configuration is 747 W [16]. Power extracted by different pattern of PSC is shown in Table 2. The PV configuration (4S2P) proposed in this paper is shown in Fig. 3.

Table 2

Different patterns of 3S1P and 4S2P configurations

4S2P-configuration 3 S 1P-configuration

Pattern 1 G11=1 kW-m-2 G21=1 kW-m-2 Pattern 3 G1= 1 kW-m-2

G12=1 kW-m-2 G22=1 kW-m-2 G2= 1 kW-m-2

G13=1 kW-m-2 G23=1 kW-m-2 G3= 1 kW-m-2

G14=1 kW-m-2 G24=1 kW-m-2

Pattern 2 G11=1 kW-m-2 G21=1 kW-m-2 Pattern 4 G1= 1 kW-m-2

G12=900 W-m-2 G22=800 W-m-2 G2= 900 W-m-2

G13=600 W-m-2 G23=700 W-m-2 G3= 900 W-m-2

G14=500 W-m-2 G24=1 kW-m-2

(1) " S

Fig. 2. Single-diode equivalent circuit of PV cell

Fig. 3. Proposed PV configuration (4S2P) 3. Methodology.

3.1. PSO algorithm. PSO is an outstanding choice for problems of nature of non-linear optimization. PSO method is inspired by the behavior of swarms, insects and flocks [17]. In the previous decade, PSO is mostly the utmost favored optimization method for MPPT applications. In PSO, the initial particles (duty cycles) are chosen randomly in domain of the bounding limits. These particles are spread in whole search domain. The best among them in each iteration is known to be P best and the finest in all repetitions is recognized as G best. Further, position and velocity of the particles/units are modified in each repetition and the

procedure keeps continuing till it reaches the best position. The motion of PSO particles in arbitrary search domain space are demonstrated in Fig. 4. The velocity and position updates are provided in:

Xt+1 = X\ + yj+1; (2)

V+1 =W y +1+Ci • ri \pl+]tl - X)+C2 • r2-(gI+1 - Xt ),(3) where W is the inertial weight; X is the position; V is the velocity; C1 and C2 are the inertia constants; t is the iteration count.

Equation (4) represents fitness function:

(4)

= aj + hj -xjj = i...n.

(5)

The opposition based optimization is defined as follow. Let q(x1, x2..., xn) is a point in «-dimensional space and q (x1 , x2 . xn ) is its opposite. For objective fitness functionf(q), iff(q') < f(q)q is kept otherwise q is kept unchanged.

3.3. PSO with OBL. PSO algorithm was presented by Eberhart R. and Kennedy J. in 1995 [21] and is based on the behavior of folk of birds. The present state of a particles that are in continues movement in the possible solution space is described by two variables, the position Xj and the velocity of movement Vj. The state of each particle in n-dimensional solution space, is determined by the position vector Xj = [xj1, xj2, ..., XjN] and velocity vector Vj = [v]1, Vfl,..., VjN]. At the start, the particles are dispersed randomly over the whole possible solution space. Its implementation in PSO is depicted in Fig. 5.

Fig.

5. Particle swarm optimization hybridized with opposition based learning

Fig. 4. Flowchart of the PSO algorithm

3.2. Opposition Based Learning (OBL). In 2005 Tizhoosh proposed the OBL technique [18] in the field of machine learning. He found that it is easier to find global maxima through opposite solution rather than random solutions and it is 50 % highly probable than original random solutions. OBL produces opposite solutions of original random solutions in each iteration and finds the fitness value of every solution and it is opposite. The one with the lower fitness is retained and other is discarded. The best among current solution space is selected and its opposite is sent to next iteration. This opposition based technique can increase the random search ability of random search methods. It has been used in PSO [19] and differential evolution algorithm [20] to improve their random search domain. The opposite point is defined as follow. Let q(x1, x2,..., xn) is a point in an «-dimensional space, x1, x2,..., x«eR, xle[al, b]. The opposite point q (x1 , x2 . xn ) of q is defined as

4. Test systems. Simulation studies for different cases of the PSO and PSO-OBL techniques have been carried out in MATLAB/Simulink R2018a. The system proposed for simulations is shown in Fig. 6. PV module is used followed by a DC-DC buck converter and load. MPPT block is consisting of an MPPT controller and PWM generator. Current and voltage from PV module are presented as input to MPPT controller which produces duty cycles accordingly to run DC-DC converter.

Fig. 6. Proposed system for PV using PSO algorithm

In Fig. 7 PV array configuration of 3S1P is shown, in which a single rail of three series connected modules in used to check partial shading effect.

x

Fig. 7. 3S1P configuration of PV panel array

Different cases of partial shading considered for 3S1P are shown in Table 2. Three different modules are given separate values of irradiance and their output P-D curve is generated which generates a 3-peaks curve corresponding to values of irradiance being given as an input. Similarly, 4S2P is shown in Fig. 8 and its different cases for partial shading are shown in Table 2.

5. Simulation results.

5.1. Simulation results of PSO and PSO-OBL for 4S2P configuration. Figures 9, 10 show simulation

Fig. 8. 4S2P configuration of PV panel array

results of 4S2P configuration for PSO-OBL in different cases of partial shading as shown in Table 2. The number of iterations used for PSO is 25 while number of iterations used for PSO-OBL is 100 which helps creating opposite candidate solutions and effectively tracks GMPP.

5.1.1. Results of pattern no. 1. In pattern no. 1 the rated power is 1992 W and the power extracted from simulation is also 1992 W. The calculated results show that tracking efficiency will be 100 % and the convergence speed will be 1.021. The P-V curve of rated power and extracted power is shown in Fig. 9.

m

V

\

\ —t—

\

1

\

* ' t, s

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.2 0.4 0.6 0.8

a b

Fig. 9. 4S2P pattern no. 1: a - characteristic curve; b - P-V curve using PSO-OBL

-

5.1.2. Results of pattern no. 2. In pattern no. 2 the rated power is 1321 W and the power extracted from simulation is 1320 W. The calculated results show that

tracking efficiency will be 99.81 % and the convergence speed will be 1.008. The P-V curve of rated power and extracted power is shown in Fig. 10.

1400 1200 1000 800 600 400 200

\

v •

r jT

/

i I t, s

Jllf

0.3 0.4

a

0.7 0.8 0

0.6 0.8 1.C

b

1.2 1.4 1.6

Fig. 10. 4S2P pattern no. 2: a - characteristic curve; b - P-V curve using PSO-OBL

5.2. Simulation results of PSO for 3S1P configuration. The parent technique (PSO) implementation for 3s1p has shown promising results as agreeing with work [22]. The characteristic curves for different patterns of 3S1P and corresponding P-V curves using PSO are shown in Fig. 11, 12.

5.2.1. Results of pattern no. 3. In pattern no. 3 the rated power is 747 W and the power extracted from simulation is also 747 W. The calculated results show that

tracking efficiency will be 100 % and the convergence speed will be 1.01. The P-V curve of rated power and extracted power is shown in Fig. 11.

5.2.2. Results of pattern no. 4. In pattern no. 4 the rated power is 688.2 W and the power extracted from simulation is 688 W. The calculated results show that tracking efficiency will be 100 % and the convergence speed will be 1.015. The P-V curve of rated power and extracted power are shown in Fig. 12.

P, W

P, W

0

t, s

1.0

1.2

0.1

0

t, s

0.1

0.2

0.5

0.6

2

...

v

A

\

y

\

\

t, s

300 200 100 0

P, W ■ -it-

_______ j-i n r r

— r

r-n___ _r-J i__1

0.4 0.5

a b

Fig. 11. 3S1P pattern no. 3: a - characteristic curve; b - P-V curve using PSO

\ .—— ^ -

[

\ t, s

' P, W

0.3

a

700 ,. 600 500 400 300 200 100 0

0.4 b

Fig. 12. 3S1P pattern no. 4: a - characteristic curve; b - P-V curve using PSO 6. Comparison among PSO and PSO-OBL. After compared to PSO because PSO doesn't perform as well as

performing simulation studies of PSO and PSO-OBL, one can see the clear improved efficiency and effective tracking of GMPP in less tin ratio (where tin is the maximum amount of time (or, more specifically, steps) that a function takes to run). This proves that PSO-OBL is better in all conditions as

PSO-OBL and especially when it comes to strong shading PSO-OBL totally outperforms PSO and proves itself to be best to use as MPPT algorithm in all environmental conditions. Tables 3, 4 provide the comparison of both the techniques.

Table 3

Comparison of both techniques

800

500

0.3

0.6

P, W

U.1

Ü.2

Ü.4

0.5

Configuration Case Algorithm t/n Power Rated power Convergence speed Oscillations at GMPP Efficiency BEST algorithm

4S2P Pattern 1 PSO 0.027 1992 1992 0.6826 0 100 PSO-OBL

PSO-OBL 0.010 1992 1.021 0 100

Pattern 2 PSO 0.027 1317.5 1321 0.6769 0 99.73 PSO-OBL

PSO-OBL 0.010 1320 1.008 0 99.81

3S1P Pattern 3 PSO 0.027 743.1 747 0.6917 0 99.72 PSO-OBL

PSO-OBL 0.010 747 1.01 0 100

Pattern 4 PSO 0.027 688 688.2 0.6811 0 99.97 PSO-OBL

PSO-OBL 0.010 688 1.015 0 99.97

Table 4

Comparison of parameters of P&O, IC, PSO and PSO-OBL

Algorithm Steady state oscillations Falling to local maximums Complexity

P&O ✓ ✓ ✓

IC ✓ ✓ ✓

PSO X X Less complex

PSO-OBL X X Less complex

7. Conclusions. Environmental conditions limit the functionality of photovoltaic module. A lot of methods have been proposed in literature to address this problem but it never achieved goal due to number of reasons. Partial shading is most known issue. Nature inspired algorithms have proven them to be good to search global maximum in a partially shaded multipeak curve which includes particle swarm optimization, artificial bee colony, flower pollination algorithm etc. Particle swarm optimization is best among these in finding global peak with less

oscillation around maximum power point, less complexity and easy to implement nature. Particle swarm optimization has disadvantage of having long computational time and converging speed particularly under strong shading conditions. In this paper, hybrid technique is used to track global max power point very efficiently by using improved particle swarm optimization algorithm. This technique is effective in all shading conditions. Simulation results shows that opposition based particle swarm optimization algorithm is less complex, more efficient, robust, flexible and have better convergence speed than particle swarm optimization and other techniques. Since shading is an unpredictable process so to predict the accuracy and to expect a 100 % output is still an uncertain thing. The increasing demand of energy and ever increasing advancement in technology is making more and more space to bring improvement to maximum power point tracking as well as the improve the transient response of the system [23, 24]. So power extraction with more accuracy in less amount of time is still a scope.

Conflict of interest. The authors declare that they have no conflicts of interest.

REFERENCES

1. Ullah M.F., Hanif A. Power quality improvement in distribution system using distribution static compensator with super twisting sliding mode control. International Transactions on Electrical Energy Systems, 2021, vol. 31, no. 9, art. no. e12997 doi: https://doi.org/10.1002/2050-7038.12997.

2. Akbar F., Mehmood T., Sadiq K., Ullah M.F. Optimization of accurate estimation of single diode solar photovoltaic parameters and extraction of maximum power point under different conditions. Electrical Engineering & Electromechanics, 2021, no. 6, pp. 46-53. doi: https://doi.org/10.20998/2074-272X.2021.6.07.

3. Hosenuzzaman M., Rahim N.A., Selvaraj J., Hasanuzzaman M., Malek A.B.M.A., Nahar A. Global prospects, progress, policies, and environmental impact of solar photovoltaic power generation. Renewable and Sustainable Energy Reviews, 2015, vol. 41, pp. 284297. doi: https://doi.org/10.1016/j.rser.2014.08.046.

4. European Photovoltaic Industry Association. Global Market Outlook for Photovoltaics 2014-2018. Brussels, May 2014. 60 p. Available at: https://helapco.gr/wp-content/uploads/EPIA Global Market Outlook for Photovolta ics 2014-2018 Medium Res.pdf (Accessed 20 August 2021).

5. Rabaia M.K.H., Abdelkareem M.A., Sayed E.T., Elsaid K., Chae K.-J., Wilberforce T., Olabi A.G. Environmental impacts of solar energy systems: A review. Science of the Total Environment, 2021, vol. 754, p. 141989. doi: https://doi.org/10.1016/i.scitotenv.2020.141989.

6. Ahmed W., Sheikh J.A., Nouman M., Ullah M.F., Mahmud M.A.P. Techno-economic analysis for the role of single end energy user in mitigating GHG emission. Energy, Sustainability and Society, 2021, vol. 11, no. 1, p. 32. doi: https://doi.org/10.1186/s13705-021-00307-3.

7. Hayder W., Abid A., Hamed M. Ben, Sbita L. MPPT based on P&O method under partially shading. 2020 17th International Multi-Conference on Systems, Signals & Devices (SSD), 2020, pp. 538-542. doi: https://doi.org/10.1109/SSD49366.2020.9364107.

8. Ali M.N., Mahmoud K., Lehtonen M., Darwish M.M.F. An Efficient Fuzzy-Logic Based Variable-Step Incremental Conductance MPPT Method for Grid-Connected PV Systems. IEEE Access, 2021, vol. 9, pp. 26420-26430. doi: https://doi.org/10.1109/ACCESS.2021.3058052.

9. Alajmi B.N., Ahmed K.H., Finney S.J., Williams B.W. Fuzzy-Logic-Control Approach of a Modified Hill-Climbing Method for Maximum Power Point in Microgrid Standalone Photovoltaic System. IEEE Transactions on Power Electronics, 2011, vol. 26, no. 4, pp. 1022-1030. doi: https://doi.org/10.1109/TPEL.2010.2090903.

10. Masoum M.A.S., Dehbonei H., Fuchs E.F. Theoretical and experimental analyses of photovoltaic systems with voltage and current-based maximum power-point tracking. IEEE Transactions on Energy Conversion, 2002, vol. 17, no. 4, pp. 514-522. doi: https://doi.org/10.1109/TEC.2002.805205.

11. Esram T., Kimball J.W., Krein P.T., Chapman P.L., Midya P. Dynamic maximum power point tracking of photovoltaic arrays using ripple correlation control. IEEE Transactions on Power Electronics, 2006, vol. 21, no. 5, pp. 1282-1291. doi: https://doi.org/10.1109/TPEL.2006.880242.

12. Kim I.-S. Sliding mode controller for the single-phase grid-connected photovoltaic system. Applied Energy, 2006, vol. 83, no. 10, pp. 1101-1115. doi: https://doi.org/10.1016/j.apenergy.2005.11.004.

13. Patel H., Agarwal V. MATLAB-Based Modeling to Study the Effects of Partial Shading on PV Array Characteristics. IEEE Transactions on Energy Conversion, 2008, vol. 23, no. 1, pp. 302-310. doi: https://doi.org/10.1109/TEC.2007.914308.

14. Ben Salah C., Ouali M. Comparison of fuzzy logic and neural network in maximum power point tracker for PV systems. Electric Power Systems Research, 2011, vol. 81, no. 1, pp. 43-50. doi: https://doi.org/10.1016/j.epsr.2010.07.005.

15. Syafaruddin, Karatepe E., Hiyama T. Artificial neural network-polar coordinated fuzzy controller based maximum power point tracking control under partially shaded conditions. IET Renewable Power Generation, 2009, vol. 3, no. 2, p. 239. doi: https://doi.org/10.1049/iet-rpg:20080065.

16. Roy Chowdhury S., Saha H. Maximum power point tracking of partially shaded solar photovoltaic arrays. Solar Energy Materials and Solar Cells, 2010, vol. 94, no. 9, pp. 1441-1447. doi: https://doi.org/10.1016/j.solmat.2010.04.011.

17. Kennedy J. Particle swarm optimization. Encyclopedia of machine learning, 2010, pp. 760-766.

18. Tizhoosh H.R. Opposition-Based Learning: A New Scheme for Machine Intelligence. International Conference on Computational Intelligence for Modelling, Control and Automation and International Conference on Intelligent Agents, Web Technologies and Internet Commerce (CMCA-IAWTIC'06), 2005, pp. 695-701, doi: https://doi.org/10.1109/CIMCA.2005.1631345.

19. Wang H., Wu Z., Rahnamayan S., Liu Y., Ventresca M. Enhancing particle swarm optimization using generalized opposition-based learning. Information Sciences, 2011, vol. 181, no.

20. pp. 4699-4714. doi: https://doi.org/10.1016/j.ins.2011.03.016.

20. Li J., Chen T., Zhang T., Li Y.X. A Cuckoo Optimization Algorithm Using Elite Opposition-Based Learning and Chaotic Disturbance. Journal of Software Engineering, 2015, vol. 10, no. 1, pp. 16-28. doi: https://doi.org/10.3923/jse.2016.16.28.

21. Poli R., Kennedy J., Blackwell T. Particle swarm optimization. Swarm Intelligence, 2007, vol. 1, no. 1, pp. 33-57. doi: https://doi.org/10.1007/s11721-007-0002-0.

22. Obukhov S., Ibrahim A., Aboelsaud R. Maximum Power Point Tracking of Partially Shading PV system Using Particle Swarm Optimization. Proceedings of the 4th International Conference on Frontiers of Educational Technologies - ICFET '18, 2018, pp. 161165. doi: https://doi.org/10.1145/3233347.3233375.

23. Anwar N., Hanif A.H., Khan H.F., Ullah M.F. Transient Stability Analysis of the IEEE-9 Bus System under Multiple Contingencies. Engineering, Technology & Applied Science Research, 2020, vol. 10, no. 4, pp. 5925-5932. doi: https://doi.org/10.48084/etasr.3273.

24. Mehdi M.F., Ahmad A., Ul Haq S.S., Saqib M., Ullah M.F. Dynamic economic emission dispatch using whale optimization algorithm for multi-objective function. Electrical Engineering & Electromechanics, 2021, no. 2, pp. 64-69. doi: https://doi.org/10.20998/2074-272X.202L2.09.

Received 09.12.2021 Accepted 23.03.2022 Published 20.04.2022

Hamdan Saeed1, MS Scholar, Tahir Mehmood1, Professor, Fadia Ali Khan2, Lecturer, Muhammad Shazaib Shah1, MS Scholar, Mian Farhan Ullah3, Lecturer, Hamid Ali1, Lecturer,

1 Department of Electrical Engineering,

University of Engineering and Technology, Taxila, Pakistan, e-mail: hamdan. saeed@students.uettaxila.edu.pk, tahir. mehmood@uettaxila.edu.pk, mshahzaib.Shah@students.uettaxila.edu.pk, engr.hamidali94@gmail.com

2 Department of Mechatronics Engineering, Wah Engineering College, University of Wah, 47040, Wah Cantt, Quaid Avenue, Pakistan, e-mail: fadia.sohail@wecuw.edu.pk

3 Department of Electrical Engineering, Wah Engineering College, University of Wah, 47040, Wah Cantt, Quaid Avenue, Pakistan,

e-mail: farhan.ullah@wecuw.edu.pk (Corresponding Author)

How to cite this article:

Saeed H., Mehmood T., Khan F.A., Shah M.S., Ullah M.F., Ali H. An improved search ability of particle swarm optimization algorithm for tracking maximum power point under shading conditions. Electrical Engineering & Electromechanics, 2022, no. 2, pp. 23-28. doi: https://doi.org/10.20998/2074-272X.2022.2.04