Interactive learning golden jackal optimization algorithm and its application in parameter extraction of proton exchange membrane fuel cells

Constructing an accurate model is crucial for designing an efficient proton exchange membrane fuel cell (PEMFC) system. However, the complexity and nonlinearity of PEMFC systems make it highly challenging to extract their parameters accurately. This paper proposes an interactive learning golden jackal optimization (ILGJO) algorithm for the precise parameter extraction of PEMFCs. The interactive learning mechanism in ILGJO effectively addresses the limitations of the traditional golden jackal optimization algorithm, such as its single evolutionary mechanism and limited information exchange among individuals. This study provides detailed publicly available experimental data on the current–voltage (I–V) characteristics of ten commercial PEMFCs and offers a comprehensive performance comparison between ILGJO and six other advanced heuristic optimization algorithms based on the same dataset. Experimental results show that the parameter extraction error of ILGJO is, on average, at least 10.96% lower than that of the six benchmark methods. Furthermore, ILGJO yields more stable parameter extraction results for PEMFCs. These findings demonstrate the superiority and effectiveness of the proposed ILGJO algorithm.