Genetic Algorithm-based Model Predictive Control for Partial Shading Mitigation on PV Array

Abstract

Photovoltaic (PV) energy harvesting systems experience performance deterioration and efficiency drops mostly due to mismatching and partial shading conditions (PSCs). Although different algorithms have been established to mitigate the negative impact due to PSCs, there are still some challenges concerning the algorithms’ robustness, accuracy, and reliability. To alleviate the effects of partial shading and enhance output power, PV array reconfiguration techniques emerged as solutions to this challenge. The purpose of this paper is to introduce the genetic algorithm (GA) based model predictive control (MPC) in order to mitigate the impact of PSCs on the PV array. This algorithm is considered due to its capability of taking multiple inputs and generating multiple output predictive signals. The proposed algorithm is implemented and simulated on MATLAB/Simulink for PV array performance optimization under PSCs. The results show that the GA-based MPC does not only mitigate the effects of PSCs but optimizes the PV array’s performance as well. From the simulation results, the GA-based MPC significantly increased the overall power output from 562.0 W (under PSCs) to 852.6 W which is almost the maximum power (under normal conditions), making an improvement of 290.6 W which is technically 51.17 % of the overall power generated by the system.