Efficient design, operation and control of commercial proton exchange membrane fuel cells (PEMFCs) in clean energy technology
Proton exchange membrane fuel cells (PEMFCs) are a promising clean energy technology with the potential to play a significant role in a sustainable energy future. Although current-voltage measurements for PEMFCs are given in manufacturer data sheets, the model parameters are unknown. The accurate an...
Gespeichert in:
| Veröffentlicht in: | International journal of hydrogen energy Jg. 197; S. 152498 |
|---|---|
| Hauptverfasser: | , |
| Format: | Journal Article |
| Sprache: | Englisch |
| Veröffentlicht: |
Elsevier Ltd
05.01.2026
|
| Schlagworte: | |
| ISSN: | 0360-3199 |
| Online-Zugang: | Volltext |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| Zusammenfassung: | Proton exchange membrane fuel cells (PEMFCs) are a promising clean energy technology with the potential to play a significant role in a sustainable energy future. Although current-voltage measurements for PEMFCs are given in manufacturer data sheets, the model parameters are unknown. The accurate and reliable identification of PEMFC model parameters is a major challenge that requires further research and advanced algorithms. Overcoming this challenge will pave the way for more efficient operation of PEMFC systems and contribute to the widespread adoption of this promising clean energy technology. With this point of view, the present study develops a novel arctic puffin optimization based on quasi-opposition-based learning and dynamic fitness-distance balance (APO-QOBL-dFDB) for more efficient design, operation, and control of PEMFC systems. The best set of seven unknown parameters (ξ1, ξ2, ξ3, ξ4, β, Rc, λ) of the Ballard Mark V, Temasek 1 kW, NedStack PS6, and BSC 500W PEMFC stacks are identified using the developed APO-QOBL-dFDB algorithm and 11 state-of-the-art metaheuristic techniques. Mean absolute error (MAE), root mean square error (RMSE), and the sum of squared error (SSE) between model predictions and experimental data are selected as objective functions. The minimum RMSE, MAE, and SSE results in 12 test cases of the PEMFC parameter identification problem were achieved by the developed APO-QOBL-dFDB algorithm. The proposed algorithm outperforms competing algorithms in 10 out of 12 PEMFC cases based on standard deviation metric results. The efficiency metric results for the APO-QOBL-dFDB are calculated to be 99.97%, 99.81%, 99.60%, and 99.94% in parameter optimization of Ballard Mark V, Temasek 1 kW, NedStack PS6, and BCS 500W PEMFC stacks, respectively. The evaluation based on the relative error (RE) metric showed that RMSE is the most suitable objective function for estimating the parameters of the examined PEMFC stacks with high accuracy. Considering all the results together, the developed APO-QOBL-dFDB algorithm comes to the fore as the best method in the PEMFC parameter identification problem with an average Friedman score of 1.1611.
•APO-QOBL-dFDB algorithm was proposed to solve PEMFC design problem.•The best set of seven unknown parameters (ξ1, ξ2, ξ3, ξ4, β, Rc, λ) were studied.•12 state-of-the-art metaheuristic algorithms used to optimize the PEMFC parameters.•The results were compared in terms of RMSE, MAE, and SSE statistical metrics.•The developed APO-QOBL-dFDB presented the best results according to the metrics. |
|---|---|
| ISSN: | 0360-3199 |
| DOI: | 10.1016/j.ijhydene.2025.152498 |