Online Modeling of a Fuel Cell System for an Energy Management Strategy Design

Energies

Loïc Boulon¹  Mohsen Kandidayeni²  Alvaro Macias²  JoãoPedro F. Trovão² 

[1] Department of Electrical & Computer Engineering, Hydrogen Research Institute, Université du Québec à Trois-Rivières, Trois-Rivières, QC G8Z 4M3, Canada;Department of Electrical & Computer Engineering, e-TESC Laboratory, University of Sherbrooke, Sherbrooke, QC J1K 2R1, Canada;

关键词: control strategy;    hybrid vehicle;    Kalman filter;    maximum power point tracker;    metaheuristic optimization;    online parameters estimation;   

DOI  :  10.3390/en13143713

来源: DOAJ

【 摘 要 】

An energy management strategy (EMS) efficiently splits the power among different sources in a hybrid fuel cell vehicle (HFCV). Most of the existing EMSs are based on static maps while a proton exchange membrane fuel cell (PEMFC) has time-varying characteristics, which can cause mismanagement in the operation of a HFCV. This paper proposes a framework for the online parameters identification of a PMEFC model while the vehicle is under operation. This identification process can be conveniently integrated into an EMS loop, regardless of the EMS type. To do so, Kalman filter (KF) is utilized to extract the parameters of a PEMFC model online. Unlike the other similar papers, special attention is given to the initialization of KF in this work. In this regard, an optimization algorithm, shuffled frog-leaping algorithm (SFLA), is employed for the initialization of the KF. The SFLA is first used offline to find the right initial values for the PEMFC model parameters using the available polarization curve. Subsequently, it tunes the covariance matrices of the KF by utilizing the initial values obtained from the first step. Finally, the tuned KF is employed online to update the parameters. The ultimate results show good accuracy and convergence improvement in the PEMFC characteristics estimation.

【 授权许可】

Unknown