Photovoltaic panel characterization by using artificial neural networks and comparison with classical models

In this paper, a characterization methodology for PV panels by using artificial neural networks (ANN) is proposed. Data typically provided by the manufacturer in the datasheet, or data obtained from environmental measurements taken at the place where the photovoltaic (PV) generation system is installed, is used for the characterization of the PV panels. The results obtained for the representation of the PV panel using ANN are presented, and then compared with results found by classical models. The results of the simulation of two PV panels using ANN show the accuracy and effectiveness of the proposed approach.