Abstract
The photovoltaic (PV) energy is essential for the future of sustainable energy developments. Conventional algorithms perform well in maximum power extraction under uniform irradiance conditions (UIC). However, they often struggle to maintain the global maximum power point (GMPP) under simple partial shading conditions (SPSCs), frequently getting stuck at local maximum power points (LMPPs) and resulting in power loss. This study developed an adapted perturb and observe based model predictive control (APO-MPC) maximum power point tracking (MPPT) approach in MATLAB/Simulink, comprising six series-connected PV modules, a boost converter, and load. The control strategy identifies GMPP and computes reference current to minimize the cost function of an optimization problem. It was compared with other MPPT algorithms regarding tracking accuracy, convergence speed, computational time, steady-state oscillations (SSOs), power efficiency under UIC, SPSCs, and complex partial shading conditions (CPSCs). The system was validated using real-time hardware implementation and seasonal field atmospheric data. The results indicated that the APO-MPC algorithm outperformed the others with no oscillations during GMPP tracking, average convergence time, and tracking efficiency of 0.17 s and 99.46%, respectively. The findings confirm its highly fast, accurate, and stable tracking of GMPP without getting trapped into LMPPs under CPSCs.