An optimized reconfiguration technique of photovoltaic array using adaptive-JAYA optimization

ABSTRACT

Reconfiguration of PV array has emerged as a promising solution to improve the power under partial shading conditions (PSC). The three main reconfiguration techniques are the physical array relocation method, electrical array reconfiguration (EAR), and the switching matrix generation based on optimization. The physical relocation techniques are complicated as they require hard labor. In the case of EAR, the optimal design of the switching matrix is still challenging. Therefore, to overcome these issues, this paper proposes a recent metaheuristic technique of Adaptive-JAYA optimization for the optimum reconfiguration of the PV array. Adaptive-JAYA is chosen for its simplicity and reliability, which makes it consume less memory, reducing the burden on processors. The proposed approaches are applied on a 9 × 9 PV array under eight shading patterns and also on a 9 × 15 PV array to analyze the competency of the proposed approach on unsymmetrical shading. The MATLAB results are quantitatively analyzed, and a comprehensive comparison is performed with many existing reconfiguration techniques. It is proved that the proposed Adaptive-JAYA approach improves the output power by 26.12%, 22.20%, 7.61%, and 7.68% compared to the TCT for the four shading patterns, i.e. SW, LW, SN, and LN.

KEYWORDS:

• PV array
• reconfiguration
• partial shading
• JAYA optimization
• maximum power extraction

Nomenclature

$I_{\mathit{ij}}$	=	Current in $i^{\mathit{th}}$ row and $j^{\mathit{th}}$ column
$G_{\mathit{ij}}$	=	Irradiation in $i^{\mathit{th}}$ row and $j^{\mathit{th}}$ column
$I_{\mathit{mij}}$	=	Maximum current in $i^{\mathit{th}}$ row and $j^{\mathit{th}}$ column
$I_{\mathit{node}}$	=	Node current
$V_a$	=	Terminal voltage
$V_{\mathit{mi}}$	=	Terminal voltage of the $i^{\mathit{th}}$ module
$X_j^i$	=	Fitness function of $j^{\mathit{th}}$ particle at $i^{\mathit{th}}$ iteration
$X_{\mathrm{best}}^i$	=	Best possible solutions
$X_{\mathrm{worst}}^i$	=	Worst possible solutions
$r_1$ and $r_2$	=	Random numbers that are distributed in the range [0,1]
$c_1^i$ and $c_2^i$	=	Adaptive coefficients
$i_{\mathit{max}}$	=	Total number of iterations
$\mathit{\alpha }$	=	Current temperature coefficient
$V_t$	=	Thermal voltage
CS	=	center shading
DS	=	diagonal shading
$\mathit{\eta }$	=	Diode ideality factor
$\mathit{TCT}$	=	Total cross tied
HC	=	Honey comb
SP	=	Series parallel
PSO	=	particle swarm optimization
HHO	=	Harris hawks optimization
AEO	=	Artificial ecosystem optimization
GWO	=	Gray wolf optimization
$V_i$and$I_i$	=	Voltage and current in the $i^{\mathit{th}}$ row respectively
$P_a$	=	Output power of the panel without considering the bypass diode
$E_e$	=	Difference between row current
$I_{\mathit{max}}$	=	Maximum current with bypass diodes
$V_m$	=	Maximum voltage
$I_m$	=	Maximum current
$V_{\mathit{oc}}$	=	Open circuit voltage
$I_{\mathit{sc}}$	=	Short circuit current
$P_{\mathit{max}}$	=	Maximum power
$P_{\mathit{avg}}$	=	Average power
LW	=	Long wide
SW	=	Short wide
LN	=	Long narrow
SN	=	Short narrow
OS	=	outer shading
RS	=	random shading
PSC	=	Partial shading conditions
GMP	=	Global maximum power
BL	=	Bridge link
EAR	=	Electrical array reconfiguration
GA	=	Genetic algorithm
BOA	=	Butterfly optimization algorithm
GOA	=	Grasshopper optimization algorithm
GSA	=	Gravitational search algorithm

Acknowledgements

This work was supported by the Science and Engineering Research Board (SERB) under the Department of Science and Technology (DST) under Grant EEQ/2022/000189.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The work was supported by the Science and Engineering Research Board [EEQ/2022/000189]