An Experimental Artificial Neural Network Based MPP Tracking for Solar Photovoltaic Systems

Abstract

This paper proposes an improved maximum power point tracking (MPPT) strategy based on Artificial Neural Network (ANN) to improve the efficiency of PV system. The proposed ANN controller tracks the MPP by estimating and adjusting the duty cycle of the DC-DC converter according to the climatic conditions (irradiance and temperature). The MPPT was trained using the measurement of the mean value of duty cycle given by a perturb and observe (P&O) algorithm under a variation of climatic data. The performances of the proposed tracking method are tested via simulation and experimental verification. The simulation results, proved that the proposed method is more efficient and can provide more energy, with less oscillation and overshoot, in comparison to the conventional P&O MPPT method. The performance of the proposed algorithm was also confirmed experimentally using dSPACE platform under relatively steady-state conditions.