Parameter estimation of PEM fuel cell by using Enhanced Arctic Puffin Optimization algorithm

• Published in: Ionics Vol. 31; no. 9; pp. 9431 - 9497
• Main Authors: Sharma, Pankaj, Raju, Saravanakumar
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.09.2025; Springer Nature B.V
• Subjects: Algorithms; Boundary conditions; Chemistry; Chemistry and Materials Science; Condensed Matter Physics; Correlation analysis; Effectiveness; Electrochemistry; Emissions; Energy Storage; Foraging behavior; Fuel cells; Genetic algorithms; Heuristic methods; Mathematical models; Optical and Electronic Materials; Optimization; Optimization algorithms; Parameter estimation; Parameter identification; Performance enhancement; Physics; Proton exchange membrane fuel cells; Renewable and Green Energy; Sensitivity analysis; Stacks; Arctic region; Arctic puffin optimization; Proton exchange membrane fuel cell; Statistical analysis; Meta-heuristic algorithms; Sum of square error
• ISSN: 0947-7047; 1862-0760
• DOI: 10.1007/s11581-025-06390-2

An essential challenge in fuel cell modelling is the identification of precise boundary conditions, often derived from the fuel cell manufacturer. In reality, not all data is provided in the manufacturer’s data sheet. Therefore, in order to enhance accuracy and estimate the performance of the cell, it is necessary to obtain all of this information. This paper presents a novel Enhanced Arctic Puffin Optimization (EnAPO) algorithm to obtain the optimal parameters of the proton exchange membrane fuel cell (PEMFC). The fitness function, subject to a set of practical constraints, is defined as the sum of squared errors (SSE). The superiority of the EnAPO algorithm was demonstrated using six distinct types of PEMFC stacks: Stack 250 W, NedStack PS6, Temasek, BCS 500-W, SR-12 500W, and Ballard Mark V PEMFC stacks and under distinct conditions (temperature and pressure). The efficacy of the proposed EnAPO algorithm is assessed by applying it to the CEC 2019, and CEC 2022 benchmark issues and subsequently comparing its results with those of existing metaheuristic (MH) algorithms under identical circumstances to demonstrate the system’s efficiency. The outcomes show that the proposed EnAPO algorithm has an SSE equal to 3.313476E - 01, 2.06556E+00, 1.23277E - 01, 1.16978E - 02, 1.42098E - 04, and 8.13912E - 01 for the Stack 250 W, NedStack PS6, Temasek, BCS 500-W, SR-12 500W, and Ballard Mark V PEMFC stacks, respectively. Furthermore, the EnAPO demonstrates enhanced performance in comparison to other MH algorithm, achieving the smallest SSE with the lowest standard deviation value (Std.). The effectiveness of the EnAPO algorithm is validated with several analyses such as convergence curve analysis, Wilcoxon’s rank-sum test and the Friedman test, boxplot study, statistical, radar plot assessment, sensitivity analysis, and correlation analysis. The final outcomes demonstrated the successful utilization of the proposed EnAPO algorithm in accurately extracting the parameters of a PEMFC model.