Parameter Extraction of Photovoltaic Cell and Module with Four Diode Model Using Flood Algorithm

Photovoltaic (PV) cells exhibit a nonlinear characteristic. Before modeling these cells, obtaining accurate parameters is essential. During the modeling phase, using these parameters is crucial for accurately characterizing and reflecting the behavior of PV structures. Therefore, this article focuses on PV parameter extraction. A PV cell and module were selected and modeled using the four-diode model (FDM). This problem, consisting of eleven unknown parameters related to the FDM, was solved with the flood algorithm (FLA). To compare the algorithm’s performance on the same problem, the polar lights optimizer (PLO), moss growth optimization (MGO), walrus optimizer (WO), and educational competition optimizer (ECO) were also employed. These five metaheuristic algorithms were used for the first time in this study, both for solving the PV parameter extraction problem and with the FDM. The objective function aimed at obtaining the smallest root mean square error (RMSE) was evaluated and compared through assessment metrics, computational accuracy, computational time, and statistical methods. The smallest minimum RMSE was obtained with FLA, calculated as 9.8251385E-04 with FDM-C and 1.6884311E-03 with FDM-M. To statistically demonstrate and reinforce FLA’s success over other algorithms, the Friedman test and Wilcoxon signed-rank test were utilized. According to these tests, FLA produced significantly better results than the other algorithms and outperformed them in pairwise comparisons. In conclusion, FLA has proven to be a successful and promising algorithm for PV parameter extraction, with its success statistically validated.