Çetinbaş İpek
Department of Electrical and Electronics Engineering Faculty of Engineering and Architecture Eskişehir Osmangazi University Eskişehir 26480 Turkey.
Glob Chall. 2024 Apr 18;8(5):2300355. doi: 10.1002/gch2.202300355. eCollection 2024 May.
This study presents the parameter extraction of single, double, and triple-diode photovoltaic (PV) models using the weighted leader search algorithm (WLS). The primary objective is to develop models that accurately reflect the characteristics of PV devices so that technical and economic benefits are maximized under all constraints. For this purpose, 24 models, 6 for two different PV cells, and 18 for six PV modules, whose experimental data are publicly available, are developed successfully. The second objective of this research is the selection of the most suitable algorithm for this problem. It is a significant challenge since the evaluation process requires using advanced statistical tools and techniques to determine the reliable selection. Therefore, seven brand-new algorithms, including WLS, the spider wasp optimizer, the shrimp and goby association search, the reversible elementary cellular automata, the fennec fox optimization, the Kepler optimization, and the rime optimization algorithms, are tested. The WLS has yielded the smallest minimum, average, RMSE, and standard deviation among those. Its superiority is also verified by Friedman and Wilcoxon signed-rank test based on 144 pairwise comparisons. In conclusion, it is demonstrated that the WLS is a superior algorithm in PV parameter extraction for developing accurate models.
本研究提出了使用加权引导搜索算法（WLS）对单二极管、双二极管和三二极管光伏（PV）模型进行参数提取。主要目标是开发能够准确反映光伏器件特性的模型，以便在所有约束条件下使技术和经济效益最大化。为此，成功开发了24个模型，其中针对两种不同光伏电池的模型有6个，针对六个光伏模块的模型有18个，其实验数据是公开可用的。本研究的第二个目标是为该问题选择最合适的算法。这是一项重大挑战，因为评估过程需要使用先进的统计工具和技术来确定可靠的选择。因此，测试了七种全新的算法，包括WLS、蜘蛛黄蜂优化算法、虾虎鱼关联搜索算法、可逆初等元胞自动机、耳廓狐优化算法、开普勒优化算法和霜优化算法。在这些算法中，WLS产生的最小值、平均值、均方根误差（RMSE）和标准差最小。基于144次成对比较的弗里德曼检验和威尔科克森符号秩检验也验证了其优越性。总之，结果表明WLS是用于开发精确模型的光伏参数提取中的一种优越算法。
