Nonlinear Neuro-Adaptive Control for MPPT Applied to Photovoltaic Systems

This paper presents an intelligent control technique using a nonlinear neuro-adaptive method. This method is based on a nonlinear model describing the dynamics of the boost converter, the PV array and the load for maximum power point tracking (MPPT) under varying environmental conditions. The proposed approach consisted of a radial basis function-neuro observer for online estimation of unknown PV system parameters (i.e., irradiation and temperature) and an online trained neuronal controller that ensures a satisfactory MPPT, whatever be the position of the photovoltaic panel. The real-time implementation of the proposed controller is achieved using Arduino Mega board. The performance of the proposed MPPT method is analyzed under different operating conditions and compared to those provided by the P&O method. Simulation results using MATLAB/Simulink software coupled to experimental results demonstrate the feasibility and the robustness of the proposed controller.