Fault diagnosis method for proton exchange membrane fuel cells based on EIS measurement optimization

• Published in: Fuel cells (Weinheim an der Bergstrasse, Germany) Vol. 22; no. 4; pp. 140 - 152
• Main Authors: Xiao, Fei, Chen, Tao, Peng, Yulin, Zhang, Rufeng
• Format: Journal Article
• Language: English
• Published: Weinheim Wiley Subscription Services, Inc 01.08.2022
• Subjects: Algorithms; Durability; EIS; Electrochemical impedance spectroscopy; Equivalent circuits; Fault detection; Fault diagnosis; Flooding; Frequencies; Fuel cells; fuzzy inference; improved KNN; Mathematical models; Optimization; Parameter identification; PEMFC; Proton exchange membrane fuel cells; Protons
• ISSN: 1615-6846; 1615-6854
• DOI: 10.1002/fuce.202200083

Poor durability and reliability are key barriers to the application of proton‐exchange membrane fuel cells (PEMFCs). The timely detection and isolation of faults can improve the performance and durability of PEMFCs. This paper proposes a PEMFC fault diagnostic method based on rapid electrochemical impedance spectroscopy (EIS) measurements. To shorten the EIS measurement time, the characteristic frequency bands were separated by a fuzzy inference method to remove interference frequency bands and low‐frequency invalid frequency bands. Then, they were optimized for the corresponding characteristic frequency band points. The parameters of the improved equivalent circuit model were identified according to the electrochemical impedance spectrum, and four of the model parameters were selected as characteristic variables for fault diagnosis. Based on this, a multifault diagnosis algorithm with an improved K‐nearest neighbor classifier applied to PEMFC was proposed. The experimental results showed that the proposed fault diagnostic method accurately and quickly distinguished four health states, that is, flooding, membrane drying, air starvation, and normal state.