Estimation of PEMFC design parameters with social learning-based optimization

Parameter estimation of proton exchange membrane fuel cell (PEMFC) design is crucial in developing and optimizing PEMFC technology. This study proposes a social learning-based swarm optimization technique to solve the PEMFC design problem. Social learning is based on particles learning from each other and sharing information in a swarm. This allows better solutions found by one particle in the swarm to be used by others, allowing them to arrive at a globally better solution in less iteration. The proposed approach uses the social learning particle swarm optimization (SL-PSO) algorithm to minimize the sum of squared errors (SSE) between the calculated and experimental voltage for the PEMFC parameter extraction problem. The study was carried out on two types of PEMFC (500 W Horizon and BCS 500 W) and compared with heuristic methods commonly used in the literature. This study is the first time in the literature that SL-PSO has been used to optimize PEMFC design parameters. The results show that the SL-PSO-based approach achieves SSE values of 3.8 × \({10}^{-2}\) ${10}^{-2}$ and 4.903 × \({10}^{-2}\) ${10}^{-2}$ for BCS 500 W and 500 W Horizon, respectively. The output reveals that the SL-PSO-based method is an effective optimization tool in PEMFC design by achieving better SSE value in minimum iteration than other comparative algorithms. The proposed approach is expected to be more helpful in making highly efficient fuel cell-based energy systems because of promising results.