Parameter Extraction of One-Diode Photovoltaic Model using Lévy Flight Directional Bat Algorithm

In this paper, a modified version of the directional bat algorithm (DBA) is proposed to fast and accurately extract the five parameters of the one-diode photovoltaic model from experimental data. In order to improve the exploit feature of the DBA, a random flight step based on the Levy distribution is introduced. Additionally a dynamic procedure to correct any solution found which violates the established parameters bounds, during the solution process, is proposed. Tests were carried out on two commercial photovoltaic devices and the results demonstrated that for these two cases at least, the proposed algorithm named Levy flight directional bat algorithm (LDBA), is more efficient and robust than the DBA and other well-established metaheuristic algorithms presented in the literature.