Estimação de parâmetros elétricos fotovoltaicos através da meta-heurística aprimorada IBFS

Abstract

Despite the metaheuristic optimization algorithm Bird Foraging Search (BFS) yielding good results when applied to the photovoltaic electrical parameter estimation problem, each of its stages could be further improved. In this work, we propose a series of modifications that lead to a derived metaheuristic, called IBFS. Its performance evaluation covered both the single-diode and two-diode models for cells and photovoltaic modules. The proposed method was tested on a diverse set of experimental I–V curves obtained under different temperature and irradiance conditions, and it was compared in terms of accuracy and dispersion with state-of-the-art reference algorithms such as SHADE, SEDE, PGJAYA, ELPSO, ITLBO, CIABC, and the original BFS. The results show that IBFS achieves high accuracy (RMSE on the order of 10−2 in the worst cases) and low dispersion (MAD on the order of 10−3 ). Compared with the other implemented algorithms, its performance was deemed significantly superior by the Wilcoxon–Mann–Whitney test in 291 out of 294 experiments, at the 5% significance level. Beyond accuracy, IBFS exhibited competitive runtimes, with only the original BFS being faster. Therefore, the contributions of this work lie in the development of enhancements to BFS and their extensive validation across a variety of real photovoltaic operating scenarios.