Proton Exchange Membrane Fuel Cell Fault Rapid Diagnosis Method Based on Electrochemical Impedance Spectroscopy and Fuzzy C-Means Algorithm

Water management is a key research direction for the performance and lifetime of proton exchange membrane fuel cell (PEMFC) stacks. The paper is aimed to develop an online fault diagnosis method that distinguishes different degrees of flooding and drying within a fuel cell stack by unobservable variables. In our research, the equivalent circuit model is established and electrochemical impedance spectroscopy (EIS) is utilized. The mathematical methods are used to extract the fault features. Fuzzy C-means is used to classify the selected features and the diagnostic rules are automatically extracted from the data. Through verification, the interpretability and computational efficiency of the proposed method are achieved.