An RBF neural network–based MPPT with sliding mode and fuzzy control for PV systems using buck converter

Abstract

This paper proposes an integrated control strategy for maximum power point tracking (MPPT) in photovoltaic (PV) systems using a buck converter. The controller combines a radial basis function (RBF) neural network for uncertainty approximation, sliding mode control (SMC) for robustness, and fuzzy logic for adaptive tuning of the switching gain to reduce chattering. The complete RBF–SMC–fuzzy control law is derived, and closed-loop stability is proven using Lyapunov theory. Simulation results in MATLAB/Simulink under both resistive and battery charging loads show that the proposed method achieves fast tracking with a settling time of about 20 ms, a tracking efficiency higher than 99%, and a voltage ripple of approximately 1.2%. Compared with conventional methods, the proposed controller significantly reduces chattering and improves power extraction performance under irradiance and load variations.