Constrained exploration method for optimal energy management in hybrid multi-stack fuel cell vehicles

Yamchi, H. B., Kandidayeni, M., Kelouwani, S. et Boulon, L. (2024). Constrained exploration method for optimal energy management in hybrid multi-stack fuel cell vehicles. Energy Conversion and Management, 316 . Article 118841. ISSN 0196-8904 DOI 10.1016/j.enconman.2024.118841

PDF Sous embargo jusqu'au 26 juillet 2026. Disponible sous licence Creative Commons Attribution Non-commercial No Derivatives. Télécharger (2MB)

Résumé

Abstract

This paper puts forward an offline optimal energy management strategy (EMS) for a multi-stack fuel cell (FC) hybrid electric vehicle (HEV) aimed at reducing both fuel consumption and degradation of FC stacks. This method is anticipated to serve as a benchmark for evaluating the performance of online EMSs in future studies. To achieve this goal, a constrained exploration method (CEM) is proposed for power allocation among different energy/power sources. Exploration methods often suffer from time-consuming processes and the curse of dimensionality, particularly when the number of control and state variables increases. To address these issues, the proposed CEM incorporates various filters for the effective calculation of battery power, battery state of charge (SOC) range, and SOC interpolation. The proposed method also utilizes a health-conscious multi-objective cost function intended for minimizing operational costs, including hydrogen consumption and degradation. CEM is introduced as a novel scalable, health-conscious and time-efficient EMS benchmark for hybrid multi-stack FC systems, efficiently decreasing operational expenses and notably shortening simulation duration. The effectiveness of the proposed method is validated through tests conducted on three heavy-duty vehicles as case studies. The first case involves an aircraft, the NASA X-57 Maxwell, fitted with two 50 kW stacks and a battery pack. The validation of the CEM is established by comparing it to level-set dynamic programming (DP) and sequential quadratic programming (SQP) in terms of accuracy and required calculation time. In the second case study, a tramway containing two 125 kW FC stacks is investigated. The validity of the CEM is approved by level-set DP and SQP algorithms. In the third case study, a truck is considered, equipped with four 75 kW FC stacks and a battery system. Level-set DP fails to do the power distribution in this case owing to the numerous state and control variables, whereas the proposed CEM successfully distributes the power among the power sources. Also, results from a comparison of the CEM and SQP methods demonstrate the superiority of CEM. This case study highlights the potential for scalability of the proposed methodology compared to the conventional DP approach found in the literature.

Type de document:	Article
Mots-clés libres:	Health-conscious energy management strategy Aeronautic Multi-stack fuel cell system Modular fuel cell system Heavy-duty vehicles Multi-objective optimization
Date de dépôt:	24 mars 2025 14:08
Dernière modification:	24 mars 2025 14:08
Version du document déposé:	Post-print (version corrigée et acceptée)
URI:	https://depot-e.uqtr.ca/id/eprint/11759

Actions (administrateurs uniquement)

Éditer la notice