Study of neural controller based MPPT in comparison with P&O for PV systems

Abstract

This study investigated the performance of two prominent maximum power point tracking (MPPT) strategies: the established perturb and observe (P&O) technique and an artificial neural network (ANN)-based controller. Through simulations conducted in MATLAB/Simulink, a 50 W photovoltaic (PV) array was evaluated under dynamic irradiance and temperature variations. Notably, data generated by the P&O system served as the training dataset for the ANN model. The simulation results indicate that the ANN controller effectively and accurately identifies the PV system’s optimal operating point even amidst fluctuating environmental conditions. When compared to the conventional P&O method, the ANN approach demonstrated superior characteristics, including a significantly faster response, diminished oscillations around the maximum power point, and enhanced tracking accuracy during rapid environmental shifts. These findings underscore the substantial potential of ANN-based MPPT strategies for improving both the efficiency and operational stability of photovoltaic power systems.