Abstract By-pass diodes are employed to mitigate hotspots in photovoltaic (PV) systems. Nonetheless, the incorporation of these diodes leads to several local maximum points in the power-voltage (P-V) characteristic of PV systems. Regardless of the prevailing environmental conditions, PV systems must function at the global maximum power point (GMPP). Classical optimization algorithms may fail to reach the GMPP in the presence of partial shading conditions (PSCs) because of their limited search capability, which can cause them to get stuck in local maximum points. Metaheuristic optimization algorithms have been shown to effectively perform Maximum Power Point Tracking (MPPT) under PSCs, thanks to their ability to explore the search space beyond local optima. In this research, a hybrid method is suggested to detect GMPP under PSCs. The proposed method combines a feedforward and backpropagation neural network (FFBPN) with a Perturb and Observe (P&O) algorithm. The primary advantages of the suggested method are as follows: 1) it performs MPPT quickly in case of PSCs, 2) the GMPP search tracking structure is simple, 3) it does not involve complex mathematical calculations, 4) minimum steady-state error, and 5) high-efficiency operation. The proposed method has been contrasted with the flower pollination algorithm (FPA), particle swarm optimization (PSO) algorithm and cuckoo search algorithm (CSA). To demonstrate the superiority of the suggested algorithm, simulation studies were conducted using MATLAB/Simulink. Additionally, the performance of the proposed method is proven by real-time application.
Yılmaz et al. (Mon,) studied this question.