A cooperative strategy-based differential evolution algorithm for robust PEM fuel cell parameter estimation

Proton exchange membrane fuel cells (PEMFCs) are powered by hydrogen energy, which is valued for its renewable, safe, and efficient characteristics, and are therefore critical in sustainable electricity generation through hydrogen electrochemical conversion. Parameter estimation in PEMFCs is a challenging but critical task, since accurate modeling is directly related to cell performance optimization and reliable energy output under different operational conditions. To improve parameter estimation accuracy, a cooperative strategy-based differential evolution (CS-DE) algorithm was developed to minimize the sum of squared errors (SSE) between experimental and simulated PEMFC voltage data for multiple BCS 500-W PEM, BCS 250-W PEM, Nedstack PS6 PEM, 500W SR-12 PEM, H-12 PEM, and HORIZON 500W PEMFC models. The CS-DE algorithm was benchmarked against standard differential evolution (DE) and other conventional methods on six commercial PEMFC types, resulting in a 15% reduction in SSE and an average improvement of 12% in estimation accuracy. These results demonstrate the robustness and adaptability of CS-DE for complex PEMFC modeling tasks.