Abstract
SummaryThe existing impedance estimation methods for proton exchange membrane fuel cell (PEMFC) based on single equivalent circuit model (ECM) cannot accurately measure electrochemical impedance spectroscopy (EIS) at all times in the future or in the whole frequency range, which affects the accuracy of state‐of‐health (SOH) estimation with EIS as aging index. Therefore, a novel health assessment approach based on multi‐model probability fusion is proposed for improving the accuracy and reliability of SOH estimation. Five typical ECMs are selected to describe PEMFC, and chaotic particle swarm optimization algorithm with dynamic inertia weight (DIW‐CPSO) is used for identifying the model parameters. The parameters related to aging are extracted for predicting the SOH, and Bayesian theorem‐based multi‐model probability approach is adopted to calculate the optimal weights of synthesizing the SOHs estimated by the five ECMs. This approach is applied to real EIS data sets of Stack FC at current 20A, 45A and 70A. The results show that the EIS estimation results of the proposed approach have better performance in the sense of root mean square error and Euclidean error. The accuracies of SOH evaluation are 46.38%, 28.69% and 10.94% higher than the single ECM respectively, and have higher accuracy compared with previous studies.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call