Precise Parameter Estimation for Polymer Electrolyte Membrane Fuel Cells Using Two Sophisticated Metaheuristic Optimization Techniques

Polymer electrolyte membrane fuel cells (PEMFCs) possess significant potential for contributing to clean energy production. The challenge of accurately modeling their polarization curves and understanding their operational characteristics has attracted great attention from researchers. This paper applies two meta-heuristic optimization techniques, namely, the dung beetle optimizer (DBO) and the rapidly exploring random tree optimization (RRTO) algorithm, to determine the unknown parameters critical for precise PEMFC modeling. The robustness of these techniques is evaluated using two different commercial PEMFC stacks under varying operating conditions. In this problem, the objective function is represented by the sum of squared errors (SSE), quantifying the discrepancy between the experimentally measured data and the outcomes produced using the estimated parameters. Also, A thorough statistical analysis incorporating various indices has been conducted to validate the robustness of the proposed approaches. A comprehensive comparison with well-known optimization strategies confirms that the DBO consistently achieves superior accuracy and convergence speed across all cases. The polarization curves obtained using DBO and RRTO closely align with the experimental data, confirming the robustness of these methods. Notably, the DBO outperforms all compared algorithms, establishing itself as the most effective PEMFC parameter estimation and optimization tool.