Parameters extraction of photovoltaic models using enhanced generalized normal distribution optimization with neighborhood search

Abstract

Abstract The photovoltaic system has been widely integrated into electrical power grids to produce clean and sustainable energy sources. Precisely modeling of systems is crucial to simulate and asset the performance of such power system. Modeling of system is a challenge because the characteristic curve of current and voltage is nonlinear and has unknown parameters due to insufficient data points in manufacture’s data sheet. This work proposes generalized normal distribution optimization based on neighborhood search strategies () to extract the parameter of single diode model (), double diode model (), and module model (). The root means square error () is used as a performance indicator. Two commercial models like RTC France solar cell and PWP201 are used to validate the ability of to precisely estimated the system’s parameters. The results show the superiority of over competitive optimization methods and can reduce the to 2.05296E-03 for PWP201 and to 9.8248E-04 for RTC France solar cell which prove that can be used as competitor method to identify the parameters of solar system. The statistical analysis shows the robustness of through statistical measurements and Wilcoxon rank test.’