Efficient and robust nonlinear control MPPT based on artificial neural network for PV system

Abstract

The objective of this paper is to optimize the energy generation of a photovoltaic system by proposing an improved maximum power point tracking (MPPT) technique. The proposed method combines an artificial neural network (ANN) with a backstepping controller to enhance the photovoltaic (PV) system’s efficiency and precision in diverse climatic conditions, including solar irradiance and temperature. The ANN is used to predict the optimal voltage at maximum power point (MPP) Vpv, ref, and the backstepping controller is used to control the DC/DC converter based on Vpv, ref. The results obtained using this technique are compared with those obtained from the perturbation and observation (P&O) technique. The proposed technique achieves better results than P&O in terms of efficiency, accuracy, stability, and response time. The simulations are performed on MATLAB/Simulink software.