An efficient ANN-based global maximum power point searching approach for solar PV system under partial shading conditions

Abstract

For solar photovoltaic (PV) systems operating in partial shading conditions (PSCs), this research suggests a simplified maximum power point tracking (MPPT) controller based on artificial neural networks (ANNs). By adding multiple local peaks (LPs) to the power-voltage (P-V) curve, PSCs make it impossible for traditional techniques to follow the global maximum power point (GMPP), which results in substantial energy loss and decreased system reliability. In order to overcome this, a DC–DC boost converter's (BC's) duty cycle is adaptively controlled by the suggested ANN-based controller for precise GMPP tracking. Three shading scenarios are used to validate the approach on a 3×2 PV array. According to simulation results, tracking efficiency is higher than 99.4% and overall system efficiency is higher than 99.3%. The ANN greatly reduces oscillations around the GMPP by achieving quick convergence (<16 ms) with little power loss (<4 W). These findings demonstrate that the suggested method is a dependable option for real-time PV applications under dynamic environmental settings, outperforming traditional MPPT techniques in terms of speed, accuracy, and stability.