Parameter extraction of photovoltaic cell/module models using starfish optimization algorithm with a secant-based objective function modification

Accurate identification of photovoltaic (PV) cell and module parameters is essential for reliable electrical modeling, performance assessment, and long-term energy yield prediction. This task is commonly formulated as an optimization problem, where the root mean square error (RMSE) between measured and estimated current-voltage characteristics is minimized. While numerous metaheuristic algorithms have been proposed to solve this problem, most existing studies focus primarily on algorithmic modifications, with limited attention given to enhancing the problem formulation itself. In this work, a recently introduced metaheuristic, the Starfish Optimization Algorithm (SFOA), is employed for PV parameter extraction and systematically evaluated against four contemporary optimization algorithms. In addition, a novel secant-based reformulation of the objective function is proposed to improve the accuracy of the parameter estimation process beyond the conventional RMSE-based approach. The proposed framework is validated on multiple PV models, including the single-diode (SDM), double-diode (DDM), and three-diode (TDM) models for PV cells, as well as the single-diode model of a PV module (PVM). Two widely used benchmark datasets, RTC France and Photowatt-PWP201, are used for experimental verification. The results demonstrate that integrating the secant-based objective function significantly enhances estimation accuracy and robustness across all considered models. In particular, the SFOA-Secant configuration achieves the lowest RMSE values of [Formula: see text] for SDM, [Formula: see text] for DDM, [Formula: see text] for TDM, and [Formula: see text] for PVM, outperforming all competing methods. These findings confirm that reformulating the objective function using the secant method constitutes an effective and complementary strategy for improving PV parameter extraction accuracy.