Design and HIL validation of improved prairie dog optimization based dynamic unitary reconfiguration for partially shaded PV arrays

This work designs an advanced dynamic unitary PV array reconfiguration scheme based on a novel improved prairie dog optimization (IPDO) to solve the power loss issue due to partial shading of PV arrays. This method takes the PV cell as the smallest entity of reconfiguration, and is suitable for large-scale PV arrays. To avoid invalid and redundant relocation of PV modules, an innovative self-rectify strategy is employed and embedded in the reconfiguration algorithm, which can reduce invalid actions of switching matrix by up to 80 %. Comparative simulation tests among nine algorithms (IPDO, genetic algorithm, particle swarm optimization, simulated annealing, four-square sudoku, arithmetic sequence pattern, modified harris hawks optimizer, artificial ecosystem-based optimization, Runge Kutta optimizer) under eight typical shading types verify the superiority of IPDO from four metrics, i.e., power loss, power enhancement, fill factor and execution ratio. Particularly, IPDO reduces power loss by 7.33 % to 30.63 %, enhances the harvested power by 6.1 % to 31.59 %, enhances fill factor (FF) by 0.579 to 0.773, and enhances execution ratio (ER) by 60.37 % to 92.67 % for asymmetric (20 x 30) PV array. For symmetric (30 x 30) PV array, IPDO reduces power loss by 9.01 % to 27.86 %, enhances the harvested power by 8.33 % to 43.69 %, enhances FF by 0.602 to 0.759, and enhances ER by 72.14 % to 90.99 %. Moreover, a new switching matrix with double-pole multi-throw switches is designed for the required switches number reduction. Lastly, hardware-in-the-loop (HIL) experiments based on RTLAB are carried out which validate the implementation feasibility of the proposed approach.