Energy Reports 7 (2021) 109–124

Contents lists available at ScienceDirect

Energy Reports
journal homepage: www.elsevier.com/locate/egyr

A robust photovoltaic array reconfiguration strategy based on coyote

optimization algorithm for enhancing the extracted power under
partial shadow condition
∗
Hegazy Rezk a,b , , Ahmed Fathy c,d , Mokhtar Aly e,f
a
College of Engineering at Wadi Addawaser, Prince Sattam Bin Abdulaziz University, Kingdom of Saudi Arabia
b
Electrical Engineering Department, Faculty of Engineering, Minia University, Egypt
c
Electrical Engineering Department, Faculty of Engineering, Jouf University, Sakaka, Kingdom of Saudi Arabia
d
Electrical Power and Machine Department, Faculty of Engineering, Zagazig University, Egypt
e
Department of Electrical Engineering, Aswan University, Aswan 81542, Egypt
f
Electronics Engineering Department, Universidad Tecnica Federico Santa Maria, Valparaiso 2390123, Chile

article info a b s t r a c t

Article history: The operation of photovoltaic (PV) array under shade faces great challenges due to the power loss
Received 19 June 2020 that reduces the extracted power. Reconfiguration process is one of the most promising solutions to
Received in revised form 28 September 2020 reduce the effect of shadow on the array. However, the physical array relocation methods like total
Accepted 7 November 2020
cross-tied (TCT) and Su Do Ku have some limitations due to their complicated arrangements. Therefore,
Available online 25 November 2020
this paper presents a recent metaheuristic approach of coyote optimization algorithm (COA) to solve
Keywords: the reconfiguration process of the partially shaded PV array. The main target is maximizing the global
Renewable energy maximum power (GMP) extracted from the array. The proposed COA is applied on 9×9 PV array
Optimization operated under four standard shadow patterns which are short wide (SW), long wide (LW), short
Shaded PV array narrow (SN), and long narrow (LN). The obtained configurations via the proposed COA method are
Global maximum power
compared to TCT, Su Do Ku, flower pollination algorithm (FPA), marine predators algorithm (MPA), and
Energy efficiency
butterfly optimization algorithm (BOA) based arrangements. The best enhancement of GMP obtained
via the proposed COA with respect to TCT configuration occurs in SW shadow pattern of 26.58%
while the least on is 7.68% placed in SN pattern. The obtained results confirmed the competence and
superiority of the proposed COA in reconfiguring the shaded array optimally.
© 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction multiple peaks in the PV array power characteristics, undesired

hot-spots, and complexities in tracking the maximum power
Global ambitious energy plans have been placed for increas- of the PV system. Therefore, extracting the maximum available
ing the renewable energy penetration in the electrical networks. power for PV sources represents a challenging task for maxi-
The solar photovoltaic (PV) generation represents one of the mizing their techno-economic and reliability benefits (Balasankar
most targeted renewable energy installations as being available et al., 2017).
everywhere, noise free, low maintenance requirements, and en- Usually, the PV panels are combined in series and/or parallel
vironmentally friendly source (te Heesen et al., 2019; Humada to form higher power PV arrays. The main widely-employed
et al., 2020). However, the dependency on environmental con- interconnection methods in the literature include the total cross-
ditions, nonlinearity property, and its irregular availability repre- tied (TCT), the honey comb (HC), the series, the series–parallel
sent the main limiting characteristics of PV generation. In partic- (SP), and the bridge link (BL) configurations (Mishra et al., 2017;
ular, the existence of dust accumulation, building shadows, and Kaushika and Gautam, 2003). In addition, other combinations
cloud passage determine the amount of generated power of PV of these techniques were presented in the literature, such as:
systems (Niazi et al., 2019; Rezk et al., 2019). These factors cause TCT-SP, TCT-HC, and TCT-BL. Another configuration based on the
partial shading (PS) condition of PV systems, which can lead to odd–even configuration (OEC) has been presented in Yadav et al.
(2020). There are two main categories for handling the existing
∗ Corresponding author at: College of Engineering at Wadi Addawaser, Prince multiple power peaks under the PS conditions. The first category
Sattam Bin Abdulaziz University, Kingdom of Saudi Arabia. is based on using soft computing methods, and the fractional
E-mail addresses: [email protected], [email protected] order (FO) controllers (Rezk et al., 2017; Al-Dhaifallah et al.,
(H. Rezk), [email protected] (A. Fathy), [email protected] (M. Aly). 2018). These methods continuously attempt to find the global

https://doi.org/10.1016/j.egyr.2020.11.035
2352-4847/© 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
H. Rezk, A. Fathy and M. Aly Energy Reports 7 (2021) 109–124

peak power point instead of solving the problem of multiple irradiance equalization is achieved through integrating dynamic
peak existence. However, the design procedures for these trackers reconfigurable portion with static portion in the same array.
include high complexities. However, this method is subjected to frequent short-circuit fault
The second category is based on the PV array reconfiguration condition due to using unbalanced switch patterns. Moreover, the
methods which have been presented to solve the multiple power performance of this method has been deteriorated at 3 × 3 PV
peak problem and to get over the complexities of the soft com- array case study in Parlak (2013). Therefore, this method is not
puting global peak searching methods. The reconfiguration pro- suitable for large-scale integrations of the PV system.
cess refers to rearrange the PV panels either physically-connected Furthermore, a method based on the image processing of
or electrically-connected so as to equalize the irradiance (Huang PV panels has been proposed in Karakose and Baygin (2014). It
et al., 2019; Pillai et al., 2018a). The reconfiguration methods are has been also shown that, effective dispersion for the shading
based on dispersing the shading from the PV panels equally so requires maintaining the average irradiances to be minimal in
as to maintain uniform row currents. In these methods, both of every row. In addition, the most dominant factors to determine
the shaded and the unshaded PV modules are highly exploited. the net outputted power are the mismatch index, the selected
The existing methods for reconfiguring the PV arrays are mainly switches, and the switching strategy among fixed and adaptive
classified into the physical array relocation (PAR) methods, and array configurations. A method has been proposed in Cheng et al.
electrical array rearrangement (EAR) methods (Sai Krishna and (2010) for the optimal selection for the switching matrix using
Moger, 2019). Fuzzy Logic (FL)-based control algorithm. However, high num-
The PAR-based methods have shown successful alternatives ber of power switching devices are needed in this method, this
to the global peak tracking methods. They are based on rear- causes increasing in its implementation cost. Another switch-
ranging the PV array one time in the planning phase (Srini- ing scheme for controlling the dynamic portion movement has
vasa Rao et al., 2014). They represent simple solutions for ef- been presented in Velasco-Quesada et al. (2009) for optimizing
fective dispersion of the shading effects. Examples of mathe- the switching of the power switches. This method has achieved
matical procedures-based physical relocation methods are the improved versatility and mobility of the PV array. However, the
Futoshiki (Sahu et al., 2016), Su Do Ku (Rani et al., 2013) and dom- employed switching scheme affects the performance of the PV
inance square (DS) (Dhanalakshmi and Rajasekar, 2018) methods. array reconfiguration. Additionally, dynamic electrical scheme
Although, long interconnection ties are required in these meth- (DES) based on inverter control has been proposed in Romano
ods for relocating the modules which increase the complexity et al. (2013) for reconfiguring the 9 × 9 PV array in grid-tied
of wiring. The Futoshiki and Su Do Ku methods face critical PV systems. In which, the number of the associated rows to
issues in reconfiguring the systems over 9 × 9 PV configura- the PV array is determined through the control algorithm of the
tion. In addition, the reconfiguration process of the DS-based inverter. However, using two different independent algorithms
method is not suitable for real-time conditions. An alternative deteriorates the flexibility of DES method.
method based on Zig-Zag method has been developed in Vijay- In dos Santos et al. (2011), a method has been presented for
alekshmy et al. (2016). This method uses particular reconfigura- reducing the required number of PV array arrangements through
tion method for row and column simultaneously for dispersing using configurations of interest (COI) parameter. However, this
the shade effects. However, this method requires complex wiring method is only suitable for 4 × 4 PV array symmetrical rearrange-
circuits, skilled labor, and additional costs. Whereas, the situa- ment. Another method based on flexible switch matrix (FSM) has
tion becomes more severe at high number of PV configurations. been proposed in Alahmad et al. (2012) for central PV inverter
A methodology based on the ancient Lo-Shu concept has been applications. The method is successful at high power extraction
proposed in Venkateswari and Rajasekar (2020) for solving the through using power-voltage (P-V) characteristics. However, the
9 × 9 PV array matrix using the submatrices of Lo-Shu concept main disadvantage of this method is sizing of asymmetrical array
with size of 3 × 3. Nine internal submatrices were included in this due to removing the shaded modules. A modified version of the
method. Additionally, magic square (Samikannu et al., 2016) and FSM has been developed in Storey et al. (2014) through separating
puzzle shade (Yadav et al., 2016) procedures have been presented the array into groups followed by dc/dc power converter stage.
for physical rearrangement of PV arrays. However, these methods The main drawbacks of this method include the complexity to
have not been validated for PV systems with higher number achieve modular level operations for particular PV array groups.
of arrays. Additionally, a method based on two-phase array has In addition, the cost of the employed system is high due to the
been presented in Pillai et al. (2018b) for reconfiguring PV arrays. high number of employed microelectronic components. Addi-
This method has straightforward initialization steps, and hence tional mathematical-based methods have been proposed in dos
higher output power is achieved. However, this method requires et al. (2015), Wang and Hsu (2011) based on using effective
complicated and tedious work around the final iteration steps for switch matrix. The well-known Newton–Raphson (NR) scheme
the array reconfiguration. has been implemented with the rough set theory for determining
From another side, in the EAR-based methods, the PV modules the switching matrix in 9 × 4 PV array system. One common
are reconfigured electrically to reach optimum shading dispersion disadvantage of the EAR-based methods is generating several
operating point. This can be achieved through continuous deter- switch pulse patterns before outputting the optimum switching
mination of the optimal reconnection of the array when there is matrix.
diagnosis of shading conditions (Yousri et al., 2020a). The EAR- Additionally, optimization methods have been presented as
based methods require controller that generates switching pulses effective solutions for avoiding the repeated switching pulses
for power electronic switches according to the desired shade dis- generation in the EAR-based methods. The intelligence and pow-
persion sequence. An adaptive EAR-based reconfiguration method erful computation of the optimization methods is employed for
for the 10 × 11 PV array configuration has been developed in generating the optimal switching scenario. Therefore, these meth-
Nguyen and Lehman (2008). A switching matrix is utilized for ods achieves prolonged lifetime of the employed power switches
achieving irradiance equalizations using the bubble sort algo- due to the reduction of the generated switching patterns. For in-
rithm. However, this method is suitable only for asymmetrical stance, the genetic algorithm (GA)-based method was presented
PV array structures. An improved method has been developed in Deshkar et al. (2015) and another method based on the particle
for the symmetrical array configuration in Liu et al. (2010). The swarm optimization (PSO) was presented in Babu et al. (2018)
method was validated through using 5 × 5 PV array structure. The for reconfiguring the PV array. In Fathy (2018), the grasshopper
110
H. Rezk, A. Fathy and M. Aly Energy Reports 7 (2021) 109–124

optimization algorithm (GOA) has been proposed for reconfigur-

ing 9×9 PV array under different shadow pattern. Yousri et al.
(2020c) introduced an approach based on artificial ecosystem op-
timizer (AEO) for reconfiguring large-scale PV of 25×25 operated
under different shadow patterns. Multi-objective optimization
problem for implementing the shaded PV array reconfiguration
has been presented in Yousri et al. (2020b) and the problem
was solved via grey wolf optimizer (GWO). The objective func-
Fig. 1. The single diode equivalent circuit of the solar cell.
tion considered in that work was minimizing the differences
in row currents and maximizing the extracted power. Ajmal
et al. (2020) reviewed all schemes assigned for shaded PV array
reconfiguration process. Additionally, the challenges faced the The current–voltage (Iph -Vph ) behavior of the PV module is
presented schemes were addressed. Fathy (2020) presented a given by Ibrahim et al. (2019):
recent approach-based butterfly optimization algorithm (BOA) to
[ ( ) ]
VPV + Rs IPV VPV + Rs IPV
reconfigure 6×4 shaded PV array optimally. IPV = Iph − Io exp −1 − (1)
Vt a Rp
However, further enhancements and developments of opti-
mization methods are highly needed for maximizing the ex- where:
tracted global maximum power (GMP) from the shaded array. IPV : The output current Io : The saturation current
Stimulated by the aforementioned issues in the PAR and EAR VPV : The output voltage Iph : The photo current
reconfiguration methods, this paper presents a new method for Vt : The thermal voltage Rs : The series resistance
reconfiguring PV arrays based on coyote optimization algorithm a: The diode ideality factor Rp : The parallel resistance
(COA). The main contributions of this paper are summarized as The solar module contains some series connected solar cells.
following: To increase the rating of PV system, groups of series and parallel
PV modules are adopted. The current–voltage (Iph -Vph ) relation of
• A new reconfiguration solution for PV array is proposed the PV array is represented by Ibrahim et al. (2019):
based on recent metaheuristic approach of coyote optimiza-
Ns
[ ( ) ]
tion algorithm (COA) for maximizing the extracted GMP in Varray + RI
Np s PV
PV systems. Iarray = Iph Np − Io Np exp Nsc
−1
Ns
Vt a
• Dynamic power enhancement method for increasing the (2)
Ns
outputted power of PV systems is proposed for mitigating Varray + RI
Np s PV
the various PS conditions. − Ns
R
Np p
• Performance validation and superiority confirmation of the
proposed COA over existing traditional solutions are pro- where:
vided at the four standard (SW, LW, SN, LN) shading sce- Varray Array voltage
narios. Iarray Array current
Nsc Number of series cells in the module
The remaining of the paper is organized as following: Section 2
Ns, N p Number of series and parallel modules, respectively
presents in details the mathematical model of PV cell, module,
and array. Different reconfiguration topologies and brief overview 3. Different reconfiguration topologies
on the coyote optimization algorithm are illustrated in Section 3
and Section 4, respectively. Section 5 highlights the proposed There are many challenges facing the operation of solar cells,
reconfiguration methodology suggested by the current research especially when they are exposed to shade. In which, they operate
work. The numerical results, analysis and discussion are pre- under the influence of shade that increases the power loss result-
sented in details in Section 6. Finally, the main finding are sum- ing in weakening of the extracted output power. Fig. 2 shows the
marized and concluded in Section 7. voltage-power and voltage–current curves for PV panel operated
under normal (fully illuminated) and shadow conditions. The
operation under shadow causes multiple local peaks of power
2. Mathematical model of PV cell, module, and array
with one global maximum power (GMP). Fig. 2 confirms that, the
GMPs are affected negatively at shade condition. Reconfiguration
Modeling of solar cell is a critical task in its proper design, process represents an effective solution to minimize the negative
which lead in turn to enhance the capability of PV system. A effects of the shadow, as the shaded panels are rearranged on
proper model of solar cell can represent the actual characteristics the same column without changing the electrical connections
of PV system. Nonetheless, modeling of solar cell is not easy of the array. Several arrangements have been constructed to
task due to the nonlinearity of PV characteristics. Accordingly, enhance the GMP extracted from the PV array operated under
many attempts have been conducted to develop a solar cell model partially shadow like series–parallel (SP), total cross tied (TCT),
with high accuracy using both analytical methods and different bridge-linked (BL), honey-comb (HC), and Su Do Ku (Woyte et al.,
optimization algorithms. There are mainly three different models 2003).
of solar cell which are single-diode model, two-diode model, and The arrangements of PV array configured via TCT and Su Do Ku
three diode model. Among the three models, the single-diode PV are shown in Fig. 3. In TCT arrangement, cross ties are employed
model is preferred due to its simplicity, easy of design, and less to connect the panels in series–parallel in each row as shown in
involved parameters. The electrical circuit diagram of a single Fig. 3(a). Each module is identified by ij, where i represents the
solar cell is shown in Fig. 1 (Mohamed et al., 2019; Abdalla et al., row number and j is the column number.
2019). The current generated by each PV model can be estimated The current produced from each panel can be calculated ac-
by applying Kirchhoff’s current law to its equivalent circuit. The cording to the subjected solar irradiance as follows:
( )
total current generated by the single-diode PV model is given as Gij
Iij = Im ij (3)
follows: Go
111
H. Rezk, A. Fathy and M. Aly Energy Reports 7 (2021) 109–124

Fig. 2. Voltage-power and voltage–current curves of PV panel under normal and shadow conditions.

Fig. 3. Traditional PV array arrangements: (a) TCT structure, (b) Su Do Ku topology.

where Im ij is the current produced from ij panel at STC. The 4. Coyote optimization algorithm
current at each node can be calculated by the aid of Kirchhoff’s
current law as follows: In this research work, a recent metaheuristic approach for
9
∑ global optimization called coyote optimization algorithm (COA) is
Iij − I(i+1)j = 0, i = 1, 2, . . . , 8
( )
Ia = (4) adopted. Its core idea is based on canis latrans species that reside
j=1 mainly north America. The algorithm is adapted to consider the
social organization of the agents called coyotes which have been
The terminal voltage of the array can be written as follows:
serving with a different algorithmic construction. The greatest
9 advantage of this method is to preserve a balance between the
∑
Va = Vmi (5) exploitation and exploration mechanisms during the optimization
i=1 process. By contrast with the grey wolf optimizer (GWO), the
where Vmi is the terminal voltage of ith panel. There are some COA does not center on how these dominant norms have been
problems in TCT as it is formed by large number of loops causing followed by these animals and the social hierarchy. Additionally,
redundancy in the array circuit. Moreover, it requires bypass the COA depends not only on the hunting preys which take
diodes that increase the power loss. Su Do Ku arrangement has place in the GWO but also on the social structure and regular
been employed to solve the problems of TCT configuration. It experience interchange which are carried out by the coyotes.
depends on changing the location of the panel in the same col- They are distinguished by cooperative functionalities as they head
umn without changing the electrical connections. However, Su towards the prey in the close chain. While, they have a strong
Do Ku arrangement has some problems as it possesses complex sense of smell which makes it possible to identify the location
connections and difficult physical relocation. of the prey. Regarding the hunting process, coyotes attack in
112
H. Rezk, A. Fathy and M. Aly Energy Reports 7 (2021) 109–124

Fig. 4. The main steps of the proposed COA optimizer.

groups, this action forces the agents to update their positions Regarding the COA algorithm, it is supposed that all the coyotes
to improve them. When coyotes striking their opponents, they have been arranged to exchange the social culture. Therefore, the
are well-prepared with a probability of threat and move away existing information of coyotes are related and calculated in term
from their current positions by an excessive random distance. The of cultural tendency (Ibrahim et al., 2019).
social behavior of the coyote serves as a set of design variables ⎧ p,t
O Nc is odd
and it is defined as follows (Pierezan and Dos Santos Coelho, ⎨ (Nc2+1)
⎪
⎪
2018): p,t
cultj = Op,t + Op,t (9)
Nc
( Nc
−
→
soccp,t = x = (x1 , x2 , x3 , . . . ., xD ) ⎩ 2 ,j +1),j
other w ise
⎪ 2
(6) ⎪
2
This social condition involves adapting the coyote to the environ- Social condition ranking of all coyotes at the t instant is presented
p,t
ment called fitc ∈ ℜ. by parameter Op,t . In other words, the median social conditions
The adaptation of the coyotes to the respective current social of all coyotes from that specific pack is employed to determine
condition is assessed as follow:
the cultural tendency of that considered pack. The birth of new
fitcp,t = f (soccp,t ) (7) coyotes, denoted agec p,t , is a function of the social combination
of two parents who are chosen randomly taking into account the
From the start of the process, the coyotes are assigned randomly
impact of the environment, it is defined as follows (Pierezan and
to the packs though the fact that they may leave sometimes their
Dos Santos Coelho, 2018);
packs and become solitary or be otherwise affiliated to another ⎧ p,t
pack. soc rndj ≺ Ps or j = j1
⎨ r1,j
⎪
⎪
The coyotes transferring between packs increases the consid- p,t p,t
erable interactions of populations with their cultures. An alpha is pupj = socr2,j rndj ≺ Ps + Pa or j = j2 (10)
⎪
chosen from the three considered COA alphas and it is described ⎪
⎩
Rj other w ise
as follows:
⏐ p,t p,t
where socr1,j and socr2,j are the social conditions of two coyotes
}
alpha = soccp,t p,t
{ ⏐
⏐ arg
⏐c ={1,2,....,Nc } min f (socc ) (8)
r1 and r2 that are selected randomly and included in the pth
113
H. Rezk, A. Fathy and M. Aly Energy Reports 7 (2021) 109–124

Fig. 5. The Pseudo code for the proposed COA method.

pack at time t. j1 and j2 are the dimensions of the optimization 5. The proposed reconfiguration methodology
problem which are selected randomly. Ps and Pa represent the
probability of scatter and the probability of association, respec- In the proposed optimization approach, the positions of solar
tively. Rj is a randomly created number in the range of variables’ radiations in the irradiance matrix are selected as coyotes, while
bounds. The cultural variety of the coyotes from the pack is the extracted GMP from the array is selected as fitness function to
conducted by the two considered probability Ps and Pa , they can be maximized. The Pseudo code of the proposed COA is given in
Fig. 5. The algorithm starts by defining the electrical specifications
be computed as follows (Pierezan and Dos Santos Coelho, 2018):
of the array and initializing packs of the solar radiations that
Ps = 1/D (11) represent the probable solutions for the problem. During the
iteration process, the coyotes (solar radiations) are assigned to
Pa = /
(1−Ps ) D (12) each pack. Then, Alpha is calculated with the aid of Eq. (8) and the
pack’s social tendency is calculated using Eq. (9). Afterwards, the
where D indicates the problem dimension inside the pack.
coyotes’ social condition is updated based on Eqs. (15) and (16),
In order to simulate such rules, two relevant parameters
the updated pack is applied to the array and the corresponding
should be considered in the design of COA. They are the worst GMP (fit pc ,t ) is recorded. Comparison between the new GMP and
fitness function w, which is represented by the solution groups, the old one is investigated with updating the solar radiation
and the number of coyotes in such group denoted Q. For better matrix in case of maximizing the GMP. Moreover, a new coyote
understanding of the processes’ rules governing, a flowchart is is generated from parents selected randomly based on Eq. (10),
shown in Fig. 4. To simulate the cultural interaction in each pack, then this new coyote is investigated to be survive or not with
on the assumption that the coyotes are just under alpha effect δ1 updating the pack information. The process is continued until the
and pack effect δ2 , which shall be determined as follows: final iteration is placed.
p,t
δ1 = alphap,t − soccr2 (13) 6. Results and discussions
p,t p,t
δ2 = cult − soccr2 (14)
The proposed methodology incorporated COA is applied on
The alpha and pack effects remain important parameters in the 9×9 PV array under four shadow patterns including short wide
update of the social coyotes’ condition as demonstrated by this (SW), long wide (LW), short narrow (SN), and long narrow (LN).
formula: Moreover, TCT and Su Do Ku arrangements for each case are in-
vestigated and compared to the obtained results via the proposed
soccp,t ,new = soccp,t ,old + r1 .δ1 + r2 .δ2 (15) COA. Furthermore, flower pollination algorithm (FPA), marine
predators algorithm (MPA), and butterfly optimization algorithm
The updating process of the social condition is achieved by taking (BOA) are programmed and employed to solve the reconfigura-
into consideration the following condition: tion process of the shaded PV array to confirm the competence
of the proposed COA. The electrical characteristics of the PV
soccp,t ,new fitcp,t ,new ≺ fitcp,t
{
module employed in this analysis are tabulated in Table 1. The
soccp,t +1 = (16)
soccp,t other w ise implemented array comprises 9 strings, each string has 9 modules
connected in series. It has a rated power of 14.113 kW during
The main steps of COA optimizer are summarized and illustrated operation under standard test conditions (G = 1000 W/m2 and
in Fig. 4. T = 25 ◦ C).
114
H. Rezk, A. Fathy and M. Aly Energy Reports 7 (2021) 109–124

Fig. 6. Architecture of shaded PV array obtained via: (a) TCT, (b) Su Do Ku, (c) FPA, (d) MPA, (e) BOA, and (d) The proposed COA in SW shadow pattern.

Table 1 as follows:
Electrical specifications of the used PV module.
Parameter Value IR1 = IR2 = IR3 = IR4 = IR5 = 9 × 0.9Im = 8.1Im
PV MPP power (W) 174.24
Open circuit voltage (V) 44.2
IR6 = 5 × 0.6Im + 4 × 0.9Im = 6.6Im
Short circuit current (A) 5.2
IR7 = IR8 = IR9 = 3 × 0.6Im + 3 × 0.4Im + 3 × 0.2Im = 3.6Im
Voltage at MPP (V) 35.2
Current at MPP (A) 4.95 (18)

The optimal configuration obtained via the proposed COA is

The population size and maximum number of iterations are shown in Fig. 6(f), in such arrangement the rows’ currents can
assigned as 50 and 100, respectively for the proposed COA and be calculated as follows:
the other programmed metaheuristic optimizers. The four studied IR1 = IR5 = IR8 = 5 × 0.9Im + 2 × 0.6Im + 0.4Im + 0.2Im = 6.3Im
shadow patterns are explained as follows:
IR2 = IR9 = 6 × 0.9Im + 2 × 0.4Im + 0.2Im = 6.4Im
Case 1: Short wide (SW) shadow pattern
IR4 = IR6 = 5 × 0.9Im + 3 × 0.6Im + 0.2Im = 6.5Im
The PV array subjected to such case of shadow can be divided
into four sub-arrays, the first one is subjected to irradiance of IR3 = IR7 = 6 × 0.9Im + 0.6Im + 0.4Im + 0.2Im = 6.6Im
900 W/m2 , while the second group receives 600 W/m2 , the third (19)
group is operated under 400 W/m2 and the last one receives
200 W/m2 . The proposed COA is implemented and the obtained The detailed calculations of the rows’ currents and powers ex-
tracted from each configuration are tabulated in Table 2. The
configuration is shown in Fig. 6 in comparison with the other
voltage–current and voltage-power curves obtained via the pro-
methods. Referring to the TCT configuration shown in Fig. 6(a),
posed COA and the others are shown in Fig. 7. The proposed COA
the current of the first row can be written as follows:
succeeded in obtaining the global maximum power of 10 405.03
W outperforming all other algorithms that were implemented.
IR1 = k11 I11 + k12 I12 + k13 I13 + · · · . + k19 I19 (17)
Case 2: Long wide (LW) shadow pattern
where k11 is the ratio between the solar radiation applied on the In this shadow patterns, the array is subjected to five different
panel located in the first row and first column of the array and the solar radiations which are 900 W/m2 , 600 W/m2 , 500 W/m2 , 400
solar radiation at STC (Go = 1000 W/m2 ). The calculated currents W/m2 , and 200 W/m2 . The TCT, Su Do Ku, FPA, MPA, BOA, and
obtained via TCT arrangement as shown in Fig. 6(a) can be written the proposed COA arrangements are shown in Fig. 8.
115
H. Rezk, A. Fathy and M. Aly Energy Reports 7 (2021) 109–124

Fig. 7. Performance comparison of the TCT, Su Do Ku, FPA, MPA, BOA, and the proposed COA arrangements in SW shadow pattern: (a) voltage–current, and (b)
voltage-power curves.

Referring to TCT configuration shown in Fig. 8(a), one can The calculations of all currents and GMPs from all studied ar-
write the rows’ currents as follows: rangements are given in Table 3, the voltage–current and voltage-
power curves for all studied arrangements are shown in Fig. 9.
IR7 = IR8 = IR9 = 2 × 0.6Im + 2 × 0.5Im + 2 × 0.4Im + 3 × 0.2Im The GMP extracted from the proposed COA arrangement in LW
= 3.6Im shadow pattern is 9044.996 W, which is the best, the next one is
9044.942 W obtained via Su Do Ku configuration. The worst one
IR1 = IR2 = IR3 = IR4 = IR5 = IR6 = 6 × 0.9Im + 3 × 0.4Im = 6.6Im
is obtained via TCT with value of 7082.865 W.
(20)
Case 3: Short narrow (SN) shadow pattern
The proposed COA arrangement gives the rows’ currents as fol- In such shadow pattern, three solar radiations strike the sur-
lows: face with values of 900 W/m2 , 600 W/m2 , and 400 W/m2 . The
obtained configuration via the proposed COA and the others are
IR3 = IR4 = 4 × 0.9Im + 0.6Im + 0.5Im + 0.4Im + 2 × 0.2Im shown in Fig. 10, the TCT arrangement shown in Fig. 10(a) results
= 5.5Im in row currents as follows:

IR1 = IR5 = IR6 = IR8 = IR9 = 4 × 0.9Im + 5 × 0.4Im = 5.6Im (21) IR8 = IR9 = 5 × 0.9Im + 4 × 0.6Im = 6.1Im
IR2 = IR7 = 4 × 0.9Im + 0.6Im + 0.5Im + 2 × 0.4Im + 0.2Im IR6 = IR7 = 5 × 0.9Im + 2 × 0.6Im + 2 × 0.4Im = 7.3Im (22)
= 5.7Im IR1 = IR2 = IR3 = IR4 = IR5 = 9 × 0.9Im = 8.1Im
116
H. Rezk, A. Fathy and M. Aly Energy Reports 7 (2021) 109–124

Table 2
GMP of TCT, Su Do Ku, FPA, MPA, BOA, and the proposed COA arrangements in SW shadow pattern.
TCT arrangement Su Do Ku arrangement FPA arrangement
Row number Current (A) Voltage (V) Power (W) Row number Current (A) Voltage (V) Power (W) Row number Current (A) Voltage (V) Power (W)
IR9 3.6Im 9Vm 32.4Vm Im IR6 6.3Im 9Vm 56.7Vm Im IR1 5.7Im 9Vm 51.3Vm Im
IR8 3.6Im – – IR7 6.3Im – – IR5 5.9Im 8Vm 47.2Vm Im
IR7 3.6Im – – IR8 6.3Im – – IR6 5.9Im – –
IR6 6.6Im 6Vm 39.6Vm Im IR1 6.3Im – – IR3 6.0Im 6Vm 36.0 Vm Im
IR5 8.1Im 5Vm 40.5Vm Im IR2 6.3Im – – IR7 6.4Im 5Vm 32.0Vm Im
IR4 8.1Im – – IR4 6.6Im 4Vm 26.4Vm Im IR4 6.8Im 4Vm 27.2Vm Im
IR2 8.1Im – – IR3 6.6Im – – IR8 6.8Im – –
IR3 8.1Im – – IR5 6.6Im – – IR2 6.9Im 2Vm 13.8Vm Im
IR1 8.1Im – – IR9 6.6Im – – IR9 7.5Im Vm 7.5Vm Im
MPA arrangement BOA arrangement The proposed COA arrangement
Row number Current (A) Voltage (V) Power (W) Row number Current (A) Voltage (V) Power (W) Row number Current (A) Voltage (V) Power (W)
IR6 6.1Im 9Vm 54.9Vm Im IR5 5.0Im 9Vm 45.0Vm Im IR1 6.3Im 9Vm 56.7Vm Im
IR3 6.3Im 8Vm 50.4Vm Im IR4 5.5Im 8Vm 44.0Vm Im IR5 6.3Im – –
IR4 6.3Im – – IR7 5.9Im 7Vm 41.3Vm Im IR8 6.3Im – –
IR5 6.3Im – – IR3 6.1Im 6Vm 36.6Vm Im IR2 6.4Im 6Vm 38.4Vm Im
IR2 6.4Im 5Vm 32.0Vm Im IR1 6.5Im 5Vm 32.5Vm Im IR9 6.4Im – –
IR8 6.4Im – – IR6 6.5Im – – IR4 6.5Im 4Vm 32.5Vm Im
IR9 6.6Im 3Vm 19.8Vm Im IR8 7.1Im 3Vm 21.3Vm Im IR6 6.5Im – –
IR7 6.7Im 2Vm 13.4Vm Im IR9 7.2Im 2Vm 14.4Vm Im IR3 6.6Im 2Vm 13.2Vm Im
IR1 6.8Im Vm 6.8Vm Im IR2 8.1Im Vm 8.1Vm Im IR7 6.6 Im – –

Fig. 8. Architecture of shaded PV array obtained via: (a) TCT, (b) Su Do Ku, (c) FPA, (d) MPA, (e) BOA, and (d) The proposed COA in LW shadow pattern.

The proposed COA arrangement results in passing row currents Table 4 shows the detailed calculations of the row currents and
GMPs obtained via all employed approaches in SN shadow pat-
as follows:
tern. The PV arrangement obtained via the proposed COA suc-
ceeded in extracting GMP of 65.7Vm Im , which is the best one
IR3 = IR5 = 7 × 0.9Im + 0.6Im + 0.4Im = 7.3Im compared to the others. The voltage–current and voltage power
IR1 = IR2 = IR6 = IR7 = IR9 = 7 × 0.9Im + 2 × 0.6Im = 7.5Im curves for the array configured via the proposed COA and the
(23) others are shown in Fig. 11.
IR2 = IR7 = 4 × 0.9Im + 0.6Im + 0.5Im + 2 × 0.4Im + 0.2Im The value of GMP obtained by the proposed COA in SN shadow
= 7.5Im pattern is 12 150.0 W while the next best one is 12 077.02 W
117
H. Rezk, A. Fathy and M. Aly Energy Reports 7 (2021) 109–124

Fig. 9. Performance comparison of the TCT, Su Do Ku, FPA, MPA, BOA, and the proposed COA arrangements in LW shadow pattern: (a) voltage–current, and (b)
voltage-power curves.

obtained via MPA. In such case the GMP is enhanced by 7.68% The row currents of the PV configuration obtained via the pro-
compared to TCT. posed COA can be calculated as follows:

Case 4: Long narrow (LN) shade pattern IR5 = IR6 = IR7 = 7 × 0.9Im + 2 × 0.4Im = 7.1Im
The last studied shadow pattern is implemented by applying IR1 = IR2 = IR3 = IR4 = IR8 = IR9 = 7 × 0.9Im + 0.7Im + 0.3Im
for solar radiations of 900 W/m2 , 700 W/m2 , 400 W/m2 , and = 7.3Im
300 W/m2 on the last three columns of the array. The proposed (25)
approach is implemented and the obtained configuration in com-
The currents of all rows and GMPs obtained by all employed
parison with the others are shown in Fig. 12. In such case, TCT
approaches are tabulated in Table 5. The obtained arrangement
configuration shown in Fig. 12(a) has row currents as follows: via COA succeeded in extracting GMP of 63.9Vm Im from the ar-
ray outperforming the others. The voltage–current and voltage-
IR8 = IR9 = 6 × 0.9Im + 3 × 0.3Im = 6.3Im power curves obtained in such case are shown in Fig. 13. The best
IR6 = IR7 = 6 × 0.9Im + 3 × 0.4Im = 6.6Im GMP is 11 746.39 W obtained by the proposed COA, the next best
(24) one is 11 684.24 obtained via Su Do Ku arrangement while the
IR3 = IR4 = IR5 = 7 × 0.9Im + 2 × 0.7Im = 7.7Im worst power is 10 843.59 W obtained via TCT. The application of
IR1 = IR2 = 9 × 0.9Im = 8.1Im COA in such case enhances the GMP by 7.69% compared to TCT.
118
H. Rezk, A. Fathy and M. Aly Energy Reports 7 (2021) 109–124

Table 3
GMP of TCT, Su Do Ku, FPA, MPA, BOA, and the proposed COA arrangements in LW shadow pattern.
TCT arrangement Su Do Ku arrangement FPA arrangement
Row number Current (A) Voltage (V) Power (W) Row number Current (A) Voltage (V) Power (W) Row number Current (A) Voltage (V) Power (W)
IR7 3.6Im 9Vm 32.9Vm Im IR2 5.5Im 9Vm 49.5Vm Im IR8 4.5Im 9Vm 40.5Vm Im
IR8 3.6Im – – IR4 5.5Im – – IR3 5.0Im 8Vm 40.0Vm Im
IR9 3.6Im – – IR3 5.6Im 7Vm 39.2Vm Im IR9 5.0Im – –
IR1 6.6Im 6Vm 39.6Vm Im IR5 5.6Im – – IR6 5.4Im 6Vm 32.4Vm Im
IR2 6.6Im – – IR6 5.6Im – – IR4 5.5Im 5Vm 27.5Vm Im
IR3 6.6Im – – IR8 5.6Im – – IR1 5.6Im 4Vm 22.4Vm Im
IR4 6.6Im – – IR9 5.6Im – – IR7 6.2Im 3Vm 18.6Vm Im
IR5 6.6Im – – IR1 5.7Im 2Vm 11.4Vm Im IR2 6.6Im 2Vm 13.2Vm Im
IR6 6.6Im – – IR7 5.7Im – – IR5 6.6Im – –
MPA arrangement BOA arrangement The proposed COA arrangement
Row number Current (A) Voltage (V) Power (W) Row number Current (A) Voltage (V) Power (W) Row number Current (A) Voltage (V) Power (W)
IR3 5.3Im 9Vm 49.5Vm Im IR3 5.3Im 9Vm 47.7Vm Im IR3 5.5Im 9Vm 49.5Vm Im
IR1 5.6Im 8Vm 44.8Vm Im IR1 5.5Im 8Vm 44.0Vm Im IR4 5.5Im – –
IR4 5.6Im – – IR4 5.6Im 7Vm 39.2Vm Im IR1 5.6Im 7Vm 39.2Vm Im
IR5 5.6Im – – IR8 5.6Im – – IR5 5.6Im – –
IR6 5.6Im – – IR9 5.6Im – – IR6 5.6Im – –
IR7 5.6Im – – IR2 5.7Im 4Vm 22.8Vm Im IR8 5.6Im – –
IR8 5.6Im – – IR5 5.7Im – – IR9 5.6Im – –
IR2 5.7Im 2Vm 11.4Vm Im IR6 5.7Im – – IR2 5.7Im 2Vm 11.4Vm Im
IR9 5.8Im Vm 5.8Vm Im IR7 5.7Im – – IR7 5.7Im – –

Fig. 10. Architecture of shaded PV array obtained via: (a) TCT, (b) Su Do Ku, (c) FPA, (d) MPA, (e) BOA, and (d) the proposed COA in SN shadow pattern.

Table 6 summarizes the values of GMPs obtained via all ap- 7. Conclusion
proaches in all studied cases. Moreover, the percentage of power
enhancement achieved with applying the proposed COA with This paper presents a new recent metaheuristic approach us-
respect to TCT based configuration is also presented. The best ing coyote optimization algorithm (COA) to solve the recon-
figuration of the partially shaded PV array. The proposed new
enhancement occurs in case 1 with 26.58% while the worst one is
reconfiguration solution for the PV array is effective in maximiz-
7.68% occurs in case 3. Fig. 14 shows the bar chart of the optimal
ing the extracted GMP in PV systems. Moreover, the proposed
GMPs of the employed approaches in all studied shadow patterns. reconfiguration method presents a simple solution. It can also
Finally, one can get that; the proposed COA is recommended as mitigate the limitations due to the complicated arrangements in
an efficient tool to enhance the extracted power from the partially the physical array relocation methods such as the total cross-
shaded PV array via solving the reconfiguration process in all tied (TCT) and Su Do Ku. The 9×9 PV array is selected as case
studied shadow patterns. study for applying the proposed reconfiguration method. The
119
H. Rezk, A. Fathy and M. Aly Energy Reports 7 (2021) 109–124

Fig. 11. Comparison of the TCT, Su Do Ku, FPA, MPA, BOA, and the proposed COA arrangements in SN shadow pattern: (a) voltage–current, and (b) voltage-power
curves.

Table 4
GMP of TCT, Su Do Ku, FPA, MPA, BOA, and the proposed COA arrangements in SN shadow pattern.
TCT arrangement Su Do Ku arrangement FPA arrangement
Row number Current (A) Voltage (V) Power (W) Row number Current (A) Voltage (V) Power (W) Row number Current (A) Voltage (V) Power (W)
IR8 6.1Im 9Vm 54.9Vm Im IR5 7.0Im 9Vm 63.0Vm Im IR1 6.8Im 9Vm 61.2Vm Im
IR9 6.1Im – – IR9 7.1Im 8Vm 56.8Vm Im IR2 7.3Im 8Vm 58.4Vm Im
IR6 7.3Im 7Vm 51.1Vm Im IR8 7.3Im 7Vm 51.1Vm Im IR5 7.3Im – –
IR7 7.3Im – – IR2 7.5Im 8Vm 60.0Vm Im IR3 7.5Im 6Vm 45Vm Im
IR5 8.1Im 5Vm 40.5Vm Im IR3 7.5Im – – IR4 7.5Im – –
IR4 8.1Im – – IR4 7.5Im – – IR6 7.5Im – –
IR3 8.1Im – – IR1 7.8Im 3Vm 23.4Vm Im IR7 7.5Im – –
IR2 8.1Im – – IR6 7.8Im – – IR8 7.8Im 2Vm 15.6Vm Im
IR1 8.1Im – – IR7 7.8Im – – IR9 8.1Im Vm 8.1Vm Im
MPA arrangement BOA arrangement The proposed COA arrangement
Row number Current (A) Voltage (V) Power (W) Row number Current (A) Voltage (V) Power (W) Row number Current (A) Voltage (V) Power (W)
IR9 7.2Im 9Vm 64.8Vm Im IR5 6.8Im 9Vm 61.2Vm Im IR3 7.3Im 9Vm 65.7Vm Im
IR2 7.3Im 8Vm 58.4Vm Im IR7 7.0Im 8Vm 56.0Vm Im IR5 7.3Im – –
IR5 7.3Im – – IR8 7.3Im – – IR1 7.5Im 7Vm 52.5Vm Im
IR8 7.3Im – – IR1 7.5Im 6Vm 45.0Vm Im IR2 7.5Im – –
IR1 7.5Im 5Vm 37.5Vm Im IR6 7.5Im – – IR6 7.5Im – –
IR6 7.5Im – – IR9 7.5Im – – IR7 7.5Im – –
IR4 7.6Im 3Vm 22.8Vm Im IR3 7.8Im 3Vm 23.4Vm Im IR9 7.5Im – –
IR3 7.8Im 2Vm 15.6Vm Im IR4 7.8Im – – IR4 7.6Im 2Vm 15.2Vm Im
IR7 7.8Im – – IR2 8.1Im Vm 8.1Vm Im IR8 7.6Im – –

120
H. Rezk, A. Fathy and M. Aly Energy Reports 7 (2021) 109–124

Fig. 12. Architecture of shaded PV array obtained via: (a) TCT, (b) Su Do Ku, (c) FPA, (d) MPA, (e) BOA, and (d) the proposed COA in LN shadow pattern.

Table 5
GMP of TCT, Su Do Ku, FPA, MPA, BOA, and the proposed COA arrangements in LN shadow pattern.
TCT arrangement Su Do Ku arrangement FPA arrangement
Row number Current (A) Voltage (V) Power (W) Row number Current (A) Voltage (V) Power (W) Row number Current (A) Voltage (V) Power (W)
IR8 6.3Im 9Vm 56.7Vm Im IR4 7.0Im 9Vm 63.0Vm Im IR7 6.3Im 9Vm 56.7Vm Im
IR9 6.3Im – – IR5 7.0Im – – IR3 6.8Im 8Vm 54.4Vm Im
IR6 6.6Im 7Vm 46.2Vm Im IR2 7.1Im 7Vm 49.7Vm Im IR6 6.8Im – –
IR7 6.6Im – – IR9 7.1Im – – IR4 7.1Im 6Vm 46.2Vm Im
IR3 7.7Im 5Vm 38.5Vm Im IR3 7.3Im 5Vm 36.5Vm Im IR8 7.3Im 5Vm 36.5Vm Im
IR4 7.7Im – – IR7 7.3Im – – IR1 7.4Im 4Vm 29.6Vm Im
IR5 7.7Im – – IR6 7.4Im 3Vm 22.2Vm Im IR2 7.6Im 3Vm 22.8Vm Im
IR1 8.1Im 2Vm 16.2Vm Im IR8 7.4Im – – IR9 7.7Im 2Vm 15.4Vm Im
IR2 8.1Im – – IR1 7.5Im Vm 7.5Vm Im IR5 8.1Im Vm 8.1Vm Im
MPA arrangement BOA arrangement The proposed COA arrangement
Row number Current (A) Voltage (V) Power (W) Row number Current (A) Voltage (V) Power (W) Row number Current (A) Voltage (V) Power (W)
IR8 6.8Im 9Vm 61.2Vm Im IR1 6.8Im 9Vm 61.2Vm Im IR5 7.1Im 9Vm 63.9Vm Im
IR1 7.0Im 8Vm 56.0Vm Im IR7 6.8Im – – IR6 7.1Im – –
IR6 7.0Im – – IR6 6.9Im 7Vm 48.3Vm Im IR7 7.1Im 7Vm 49.7Vm Im
IR5 7.2Im 6Vm 43.2Vm Im IR5 7.0Im 6Vm 42.0Vm Im IR1 7.3Im – –
IR3 7.3Im 5Vm 36.5Vm Im IR3 7.3Im 5Vm 36.5Vm Im IR2 7.3Im – –
IR4 7.4Im 4Vm 29.6Vm Im IR8 7.3Im – – IR3 7.3Im – –
IR9 7.4Im – – IR9 7.3Im – – IR4 7.3Im – –
IR2 7.5Im 2Vm 15.0Vm Im IR2 7.6Im 2Vm 15.2Vm Im IR8 7.3Im 2Vm 14.6Vm Im
IR7 7.5Im – – IR4 8.1Im Vm 8.1Vm Im IR9 7.3Im – –

Table 6
Summary of the optimal GMPs obtained via COA and the other approaches.
Case # GMP (W) Power enhancement
(w.r.t TCT)
TCT arrangement Su Do Ku FPA arrangement MPA arrangement BOA arrangement The proposed
arrangement COA
arrangement
1 7639.058 10 397.96 9804.605 10 301.21 8810.686 10 405.03 26.58%
2 7082.865 9 044.942 7860.267 8944.046 8942.743 9044.996 21.69%
3 11 217.45 11 920.51 11 737 12 077.02 11 712.07 12 150.9 7.68%
4 10 843.59 11 684.24 10 970.19 11 563.18 11 477.67 11 746.39 7.69%

121
H. Rezk, A. Fathy and M. Aly Energy Reports 7 (2021) 109–124

Fig. 13. Comparison of TCT, Su Do Ku, FPA, MPA, BOA, and the proposed COA arrangements in LN shadow pattern: (a) voltage–current, and (b) voltage-power curves.

Fig. 14. Bar chart of GMPs obtained via all employed approaches in all studied cases.

122
H. Rezk, A. Fathy and M. Aly Energy Reports 7 (2021) 109–124

results confirm the superiority of the proposed COA reconfigu- dos, P., Vicente, S., Pimenta, T.C., Ribeiro, E.R., 2015. Photovoltaic array reconfig-
ration method at the four standard shade scenarios (SW, LW, uration strategy for maximization of energy production. Int. J. Photoenergy
2015, 1–11. http://dx.doi.org/10.1155/2015/592383.
SN, LN) compared to the TCT, Su Do Ku, FPA, MPA, and BOA-
dos Santos, P., Vicente, E.M., Ribeiro, E.R., 2011. Relationship between the
based arrangements. The results show that, the proposed COA shading position and the output power of a photovoltaic panel. In: XI
reconfiguration method enhances the extracted GMP over the Brazilian Power Electronics Conference. Natal, Brazil. pp. 676–681. http:
traditional TCT method by 26.58%, 21.69%, 7.68%, and 7.69% at //dx.doi.org/10.1109/COBEP.2011.6085183.
the SW, LW, SN, and LN share scenarios, respectively. Therefore, Fathy, A., 2018. Recent meta-heuristic grasshopper optimization algorithm for
optimal reconfiguration of partially shaded PV array. Sol. Energy 171,
the obtained results confirmed the competence and superiority
638–651. http://dx.doi.org/10.1016/j.solener.2018.07.014.
of the proposed COA in reconfiguring the shaded PV array op- Fathy, A., 2020. Butterfly optimization algorithm based methodology for en-
timally. The author recommended the application of COA based hancing the shaded photovoltaic array extracted power via reconfiguration
approach experimentally as the obtained results confirmed its process. Energy Convers. Manage. 220, 113115. http://dx.doi.org/10.1016/j.
competence in solving the reconfiguration process of shaded PV enconman.2020.113115.
Huang, Y.-P., Chen, X., Ye, C.-E., 2019. Implementation of a modified circuit
array. Moreover, the dynamic cases of cloud movement and dust reconfiguration strategy in high concentration photovoltaic modules under
accumulative will be considered in the future works that the partial shading conditions. Sol. Energy 194, 628–648. http://dx.doi.org/10.
authors intend to implement in the future 1016/j.solener.2019.10.038.
Humada, A.M., et al., 2020. Modeling of PV system and parameter extraction
based on experimental data: Review and investigation. Sol. Energy 199,
CRediT authorship contribution statement
742–760. http://dx.doi.org/10.1016/j.solener.2020.02.068.
Ibrahim, M.N., Rezk, H., Al-Dhaifallah, M., Sergeant, P., 2019. Solar array fed
Hegazy Rezk: Conceptualization, Formal analysis, Investiga- synchronous reluctance motor driven water pump: an improved perfor-
tion, Methodology, Software, Writing - original draft, Writing - mance under partial shading conditions. IEEE Access 7, 77100–77115. http:
review & editing. Ahmed Fathy: Conceptualization, Investigation, //dx.doi.org/10.1109/ACCESS.2019.2922358.
Karakose, M., Baygin, M., 2014. Image processing based analysis of moving
Methodology, Software, Writing - original draft, Writing - review
shadow effects for reconfiguration in PV arrays. In: 2014 IEEE International
& editing. Mokhtar Aly: Formal analysis, Methodology, Writing - Energy Conference (ENERGYCON). Cavtat, Croatia. pp. 683–687. http://dx.doi.
original draft, Writing - review & editing. org/10.1109/ENERGYCON.2014.6850500.
Kaushika, N.D., Gautam, N.K., 2003. Energy yield simulations of interconnected
Declaration of competing interest solar pv arrays. IEEE Trans. Energy Convers. 18 (1), 127–134. http://dx.doi.
org/10.1109/TEC.2002.805204.
Liu, Yanli, Pang, Zhichao, Cheng, Ze, 2010. Research on an adaptive solar photo-
The authors declare that they have no known competing finan- voltaic array using shading degree model-based reconfiguration algorithm.
cial interests or personal relationships that could have appeared In: 2010 Chinese Control and Decision Conference. Xuzhou, China. pp.
to influence the work reported in this paper. 2356–2360. http://dx.doi.org/10.1109/CCDC.2010.5498823.
Mishra, N., Yadav, A.S., Pachauri, R., Chauhan, Y.K., Yadav, V.K., 2017. Performance
enhancement of PV system using proposed array topologies under vari-
Acknowledgment ous shadow patterns. Sol. Energy 157, 641–656. http://dx.doi.org/10.1016/
j.solener.2017.08.021.
This project was supported by the Deanship of Scientific Re- Mohamed, Mohamed A., Zaki Diab, Ahmed A., Rezk, Hegazy, 2019. Partial shading
search at Prince Sattam Bin Abdulaziz University, Kingdom of mitigation of PV systems via different meta-heuristic techniques. Renew.
Saudi Arabia under the research project No 2020/01/11742. Energy 130, 1159–1175.
Nguyen, D., Lehman, B., 2008. An adaptive solar photovoltaic array using model-
based reconfiguration algorithm. IEEE Trans. Ind. Electron. 55 (7), 2644–2654.
References http://dx.doi.org/10.1109/TIE.2008.924169.
Niazi, K.A.K., Yang, Y., Sera, D., 2019. Review of mismatch mitigation techniques
Abdalla, Omer, Rezk, Hegazy, Ahmed, Emad M., 2019. Wind driven optimiza- for PV modules. IET Renew. Power Gener. 13 (12), 2035–2050. http://dx.doi.
tion algorithm based global MPPT for PV system under non-uniform solar org/10.1049/iet-rpg.2019.0153.
irradiance. Sol. Energy 180, 429–444. Parlak, K.S., 2013. A New reconfiguration method for PV array system. In:
Ajmal, A.M., Babu, T.S., Ramachandaramurthy, V.K., Yousri, D., Ekanayake, J.B., IECON 2013-39th Annual Conference of the IEEE Industrial Electronics So-
2020. Static and dynamic reconfiguration approaches for mitigation of partial ciety. Vienna, Austria. pp. 1478–1483. http://dx.doi.org/10.1109/IECON.2013.
shading influence in photovoltaic arrays. Sustain. Energy Technol. Assess. 40, 6699351.
100738. http://dx.doi.org/10.1016/j.seta.2020.100738. Pierezan, J., Dos Santos Coelho, L., 2018. Coyote optimization algorithm: a new
Al-Dhaifallah, M., Nassef, A.M., Rezk, H., Nisar, K.S., 2018. Optimal parameter metaheuristic for global optimization problems. In: 2018 IEEE Congress on
design of fractional order control based INC-MPPT for PV system. Sol. Energy Evolutionary Computation (CEC). Rio de Janeiro. pp. 1–8. http://dx.doi.org/
159, 650–664. http://dx.doi.org/10.1016/j.solener.2017.11.040. 10.1109/CEC.2018.8477769.
Alahmad, M., Chaaban, M.A., kit Lau, S., Shi, J., Neal, J., 2012. An adaptive Pillai, D.S., Prasanth Ram, J., Siva Sai Nihanth, M., Rajasekar, N., 2018a. A
utility interactive photovoltaic system based on a flexible switch matrix simple, sensorless and fixed reconfiguration scheme for maximum power
to optimize performance in real-time. Sol. Energy 86 (3), 951–963. http: enhancement in PV systems. Energy Convers. Manage. 172, 402–417. http:
//dx.doi.org/10.1016/j.solener.2011.12.028. //dx.doi.org/10.1016/j.enconman.2018.07.016.
Babu, T.S., Ram, J.P., Dragicevic, T., Miyatake, M., Blaabjerg, F., Rajasekar, N., Pillai, D.S., Rajasekar, N., Ram, J.P., Chinnaiyan, V.K., 2018b. Design and testing
2018. Particle swarm optimization based solar PV array reconfiguration of of two phase array reconfiguration procedure for maximizing power in solar
the maximum power extraction under partial shading conditions. IEEE Trans. PV systems under partial shade conditions (PSC). Energy Convers. Manage.
Sustain. Energy 9 (1), 74–85. http://dx.doi.org/10.1109/TSTE.2017.2714905. 178, 92–110. http://dx.doi.org/10.1016/j.enconman.2018.10.020.
Balasankar, R., Arasu, G.T., Christy Mano Raj, J.S., 2017. A global MPPT technique Rani, B.I., Ilango, G.S., Nagamani, C., 2013. Enhanced power generation from PV
invoking partitioned estimation and strategic deployment of P & O to tackle array under partial shading conditions by shade dispersion using Su Do Ku
partial shading conditions. Sol. Energy 143, 73–85. http://dx.doi.org/10.1016/ configuration. IEEE Trans. Sustain. Energy 4 (3), 594–601. http://dx.doi.org/
j.solener.2016.12.018. 10.1109/TSTE.2012.2230033.
Cheng, Ze, Pang, Zhichao, Liu, Yanli, Xue, Peng, 2010. An adaptive solar photo- Rezk, H., Aly, M., Al-Dhaifallah, M., Shoyama, M., 2019. Design and Hardware
voltaic array reconfiguration method based on fuzzy control. In: 2010 8th Implementation of New Adaptive fuzzy logic-based MPPT control method for
World Congress on Intelligent Control and Automation. Jinan, China. pp. Photovoltaic applications. IEEE Access 1. http://dx.doi.org/10.1109/ACCESS.
176–181. http://dx.doi.org/10.1109/WCICA.2010.5553911. 2019.2932694.
Deshkar, S.N., Dhale, S.B., Mukherjee, J.S., Babu, T.S., Rajasekar, N., 2015. Solar PV Rezk, H., Fathy, A., Abdelaziz, A.Y., 2017. A comparison of different global
array reconfiguration under partial shading conditions for maximum power MPPT techniques based on meta-heuristic algorithms for photovoltaic system
extraction using genetic algorithm. Renew. Sustain. Energy Rev. 43, 102–110. subjected to partial shading conditions. Renew. Sustain. Energy Rev. 74,
http://dx.doi.org/10.1016/j.rser.2014.10.098. 377–386. http://dx.doi.org/10.1016/j.rser.2017.02.051.
Dhanalakshmi, B., Rajasekar, N., 2018. Dominance square based array reconfigu- Romano, P., Candela, R., Cardinale, M., Li Vigni, V., Musso, D., Riva Sanseverino, E.,
ration scheme for power loss reduction in solar PhotoVoltaic (PV) systems. 2013. Optimization of photovoltaic energy production through an efficient
Energy Convers. Manage. 156, 84–102. http://dx.doi.org/10.1016/j.enconman. switching matrix. J. Sustain. Dev. Energy Water Environ. Syst. 1 (3), 227–236.
2017.10.080. http://dx.doi.org/10.13044/j.sdewes.2013.01.0017.

123
H. Rezk, A. Fathy and M. Aly Energy Reports 7 (2021) 109–124

Sahu, H.S., Nayak, S.K., Mishra, S., 2016. Maximizing the power generation of Vijayalekshmy, S., Bindu, G.R., Rama Iyer, S., 2016. A novel Zig-Zag scheme for
a partially shaded PV array. IEEE J. Emerg. Sel. Top. Power Electron. 4 (2), power enhancement of partially shaded solar arrays. Sol. Energy 135, 92–102.
626–637. http://dx.doi.org/10.1109/JESTPE.2015.2498282. http://dx.doi.org/10.1016/j.solener.2016.05.045.
Sai Krishna, G., Moger, T., 2019. Reconfiguration strategies for reducing partial Wang, Y.-J., Hsu, P.-C., 2011. An investigation on partial shading of PV mod-
shading effects in photovoltaic arrays: State of the art. Sol. Energy 182, ules with different connection configurations of PV cells. Energy 36 (5),
429–452. http://dx.doi.org/10.1016/j.solener.2019.02.057. 3069–3078. http://dx.doi.org/10.1016/j.energy.2011.02.052.
Samikannu, S.M., Namani, R., Subramaniam, S.K., 2016. Power enhancement of Woyte, A., Nijs, J., Belmans, R., 2003. Partial shadowing of photovoltaic arrays
partially shaded PV arrays through shade dispersion using magic square with different system configurations: literature review and field test re-
configuration. J. Renew. Sustain. Energy 8 (6), 063503. http://dx.doi.org/10. sults. Sol. Energy 74 (3), 217–233. http://dx.doi.org/10.1016/S0038-092X(03)
1063/1.4972285. 00155-5.
Srinivasa Rao, P., Saravana Ilango, G., Nagamani, C., 2014. Maximum power from Yadav, K., Kumar, B., Swaroop, D., 2020. Mitigation of mismatch power losses of
PV arrays using a fixed configuration under different shading conditions. PV array under partial shading condition using novel odd even configuration.
IEEE J. Photovolt. 4 (2), 679–686. http://dx.doi.org/10.1109/JPHOTOV.2014. Energy Rep. 6, 427–437. http://dx.doi.org/10.1016/j.egyr.2020.01.012.
2300239. Yadav, A.S., Pachauri, R.K., Chauhan, Y.K., 2016. Comprehensive investigation
Storey, J., Wilson, P.R., Bagnall, D., 2014. The optimized-string dynamic photo- of PV arrays with puzzle shade dispersion for improved performance. Sol.
voltaic array. IEEE Trans. Power Electron. 29 (4), 1768–1776. http://dx.doi. Energy 129, 256–285. http://dx.doi.org/10.1016/j.solener.2016.01.056.
org/10.1109/TPEL.2013.2265497. Yousri, D., Allam, D., Eteiba, Magdy.B., 2020a. Optimal photovoltaic array recon-
te Heesen, H., Herbort, V., Rumpler, M., 2019. Performance of roof-top PV figuration for alleviating the partial shading influence based on a modified
systems in Germany from 2012 to 2018. Sol. Energy 194, 128–135. http: harris hawks optimizer. Energy Convers. Manage. 206, 112470. http://dx.doi.
//dx.doi.org/10.1016/j.solener.2019.10.019. org/10.1016/j.enconman.2020.112470.
Velasco-Quesada, G., Guinjoan-Gispert, F., Pique-Lopez, R., Roman-Lumbreras, M., Yousri, D., Babu, T.S., Balasubramanian, K., Osama, A., Fathy, A., 2020b. Multi-
Conesa-Roca, A., 2009. Electrical PV array reconfiguration strategy for energy objective grey wolf optimizer for optimal design of switching matrix for
extraction improvement in grid-connected PV systems. IEEE Trans. Ind. shaded PV array dynamic reconfiguration. IEEE Access http://dx.doi.org/10.
Electron. 56 (11), 4319–4331. http://dx.doi.org/10.1109/TIE.2009.2024664. 1109/ACCESS.2020.3018722.
Venkateswari, R., Rajasekar, N., 2020. Power enhancement of PV system via Yousri, D., Babu, T.S., Mirjalili, S., Rajasekar, N., Abd Elaziz, M., 2020c. A novel ob-
physical array reconfiguration based Lo Shu technique. Energy Convers. jective function with artificial ecosystem-based optimization for relieving the
Manage. 215, 112885. http://dx.doi.org/10.1016/j.enconman.2020.112885. mismatching power loss of large-scale photovoltaic array. Energy Convers.
Manage. 225, 113385. http://dx.doi.org/10.1016/j.enconman.2020.113385.

124