Experimental Assessment of MPPT Based on a Neural Network Controller

Improving the efficiency of photovoltaic systems depends mainly on extracting the maximum power output from the panels with high accuracy in tracking the maximum power point (MPP) despite variations in load and weather conditions (temperature and solar irradiance). In this paper, a neural network (ANN) based maximum power point tracking (MPPT) controller has been developed and experimentally validated under real conditions. The objective is to overcome the drawbacks of the classical MPPT algorithms (P&O and INC), namely the response time and the oscillation around the MPP. The obtained results show a high accuracy of the MPP tracking.