Multiple Synergistic Degradation Parameter Identification of PEM Fuel Cells Utilizing Threat Response Adaptive Differential Evolution Algorithm

Proton exchange membrane fuel cells (PEMFCs) experience significant performance degradation over long-term operation, hindering their commercial viability. Accurately identifying polarization curve parameters during aging is crucial for elucidating degradation mechanisms and enabling health monitoring, yet this task faces challenges such as parametric coupling and pronounced nonlinearity. This study tackles these identification challenges through the integrated application of a dynamic aging model, which captures the synergy between degradation mechanisms like platinum oxidation and membrane resistance increase, and the introduction of the novel Threat Response Adaptive Differential Evolution (TRADE) algorithm. The algorithm employs multidimensional threat assessment, a three-tier response strategy, and adaptive decision-making to achieve accurate and robust parameter identification. Validated with experimental data from commercial PEMFC stacks over a full ageing cycle, the TRADE algorithm achieves a root mean square error as low as 0.00675 V within the 100–1000 h range, demonstrating superior fitting performance and stability. Sensitivity analysis further reveals that activation overpotential is the dominant degradation mechanism throughout the entire cycle (contributing ≥ 70%), whereas the contribution of concentration overpotential rises substantially to 33% under high-current-density conditions. This study provides a robust modelling framework and an effective methodology for quantifying PEMFC ageing mechanisms, predicting remaining useful life, and optimizing system performance.