An Optimized Shade-Tolerant MPPT for Photovoltaic Systems Using AI-Classification

Multiple artificial intelligence (AI) regression and metaheuristic search techniques have been proposed for Maximum Power Point Tracking (MPPT) of Photovoltaic (PV) systems. However, their reliability under Partial Shading (PS) is still questionable given that it occurs at various depths, shapes, and surface areas. This work presents a robust Maximum Power Point Tracking (MPPT) technique for Photovoltaic (PV) systems under Partial Shading (PS). Motivated by the ON/OFF nature of the bypass diodes, the infinite PS patterns are grouped into distinct discretized regions each containing a single MPP. An artificial neural classifier is trained to identify the region of the Global MPP (GMPP). Once the GMPP class is obtained, a Grey Wolf Optimization (GWO) is launched to locate it. Additionally, a class zero is assigned to account for the overlapping of the regions due to the expansion and contraction of the I-V curve for extreme temperature and irradiance conditions. The superiority of the proposed method is verified experimentally.