Improved Artificial Neural Network Based MPPT Tracker for PV System under Rapid Varying Atmospheric Conditions

Abstract

The main role of maximum power point tracker (MPPT) is to adapt the optimal resistance RMPP , corresponding to the maximum power point (MPP) of the photovoltaic generator (GPV), to the impedance of the load for maximum power transfer. This is accomplished through the tuning of the duty cycle D to an optimum value DMPP , that controls a DC-DC converter applied between the GPV and the load Rload . This paper proposes a system that is applicable to any load and enables rapid and precise tracking under variable weather circumstances. The suggested scheme allows simple and direct computation of the control signal DMPP from the values of Rload and RMPP . Rload is computed using two voltage and current sensors, while RMPP is estimated using an artificial neural network (ANN) that employs the solar irradiance, temperature and the GPV internal current-voltage characteristics. Using MATLAB environment, the obtained simulation results reveal better and more effective tracking with nearly no oscillations compared to a relevant ANN-based technique, under various meteorological conditions.