ANN-Based MPPT Algorithm for Photovoltaic Systems

Abstract

It is very important to get maximum efficiency from photovoltaic panels with low yields. To be able to achieve high efficiency from panels, maximum power point tracking algorithms have been developed. Perturb&Observe and incremental conductance methods, which are among the conventional methods, are not very successful in capturing the points from which maximum power can be obtained in variable atmospheric conditions. In this article, a maximum power point tracking method based on the artificial neural network was proposed. In the proposed method, artificial neural network inputs were designed as temperature and voltage, while its output was designed as the reference voltage. By controlling this reference voltage through a PI controller, it was ensured that the system generated maximum power in variable atmospheric conditions. Conventional methods and the proposed method were compared by simulation studies conducted in the MATLAB/Simulink environment. The superiority of the proposed method was demonstrated with a compelling scenario in which temperature and radiation were constantly changing.