Optimal PEM Fuel Cell Model Using a Novel Circle Search Algorithm

The aim of this article is to introduce a novel Circle Search Algorithm (CSA) with the purpose obtaining precise electrical model proton exchange membrane fuel cell (PEMFC). Current-voltage and current-power curves are used characterize performance PEMFCs. A nonlinear seven unknown parameters describe these polarization curves. Estimating critical issue because they influence dynamic analysis cells in variety applications such as transportation smart grids. suggested method based on minimizing fitness function (the sum squared errors (SSE)) between estimated measured voltage values. CSA compared neural network algorithm (NNA), grey wolf optimization (GWO), sine cosine (SCA). results reveal that simulation times CSA, NNA, GWO, SCA 5.2, 6, 5.8, 5.75 s, respectively. Moreover, converges best minimum within first 100 iterations, which faster than other algorithms. robustness verified using 20 independent runs, where achieves smallest average standard deviation. In addition, t-test proves superiority algorithms, all p-values less 5%. simulated I-V I-P CSA-PEMFC match very closely. efficacy evaluated under temperature pressure conditions. Therefore, has potential be an accurate efficient model.