Improved Metaheuristic Algorithms for Optimal Parameters Selection of Proton Exchange Membrane Fuel Cells: A Comparative Study

• Published in: IEEE access Vol. 11; pp. 7369 - 7397
• Main Authors: Abdel-Basset, Mohamed, Mohamed, Reda, Abdel-Fatah, Laila, Sharawi, Marwa, Sallam, Karam M.
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 2023; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Cathodes; Comparative studies; Fuel cells; Heuristic methods; Mathematical models; metaheuristic algorithms; Metaheuristics; modeling; Optimization; Optimization algorithms; Optimization methods; Parameters; PEMFC; Performance measurement; Proton exchange membrane fuel cells; Protons; self-adaptive strategy; Source code; Stacks; Steady-state; steady-state characterization; Voltage
• ISSN: 2169-3536; 2169-3536
• DOI: 10.1109/ACCESS.2023.3236023

As a new attempt to design a precise mathematical model for the proton exchange membrane fuel cell (PEMFC), in this paper, three recently-proposed well-established optimizers: horse herding optimization algorithm (HOA), seagull optimization algorithm (SOA) and gradient-based optimizer (GBO) integrated with two newly-proposed effective strategies, namely self-adaptive strategy, and ranking-based updating strategy, have been extensively investigated to accurately estimate the unknown parameters of this model for accomplishing a better output voltage of the simulated PEMFC stacks. Those hybridized algorithms were briefly named HHOA, HSOA, and HGBO. To assess the performance of those proposed algorithms, six common PEMFC stacks were used and their outcomes were extensively compared with the standard algorithms and some of the state-of-the-arts under various performance metrics and the Wilcoxon rank-sum test. The experimental findings show the effectiveness of both HGBO and HSOA in terms of convergence speed and final accuracy; However, HSOA could be more stable. The source code of this study is publicly available at https://drive.matlab.com/sharing/d9263036-9f80-4a40-bad9-ad476ed19c69 .