Performance evaluation of conventional and ANN-based MPPT algorithms for PV systems in dynamic environmental conditions

This paper presents a rigorous performance comparison of five maximum power point tracking (MPPT) algorithms for photovoltaic (PV) systems under environmental conditions using MATLAB/Simulink. The methods evaluated include perturb and observe (P&O), incremental conductance (INC), fractional open circuit voltage (FOCV), fractional short circuit current (FSCC), a modern optimization-based grey wolf optimizer (GWO), and an artificial neural network (ANN)-based controller. Conventional MPPT methods often show reduced performance, experiencing delayed tracking, increased oscillations, and significant energy loss, particularly during rapid variations in irradiance and temperature. To address these limitations, a predictive ANN-based MPPT controller is developed using a dynamically generated dataset derived from varying irradiance and temperature profiles. Unlike conventional iterative methods, the proposed ANN directly estimates the optimal operating voltage (V MPP ), enabling faster convergence and improved stability. Simulation results demonstrate that the ANN-based controller achieves the highest efficiency (97.55%), the fastest transient response (20–30 ms), and minimal steady-state power fluctuation (<5 W) among all evaluated MPPT techniques. The GWO-based method also shows improved performance over conventional techniques but remains slightly inferior to the ANN-based approach. These results confirm that the proposed ANN-based MPPT controller provides a robust and efficient solution for PV systems operating under rapidly changing environmental conditions.