A Fuzzy Logic-based Method for Proton Exchange Membrane Fuel Cell Fault Diagnosis

Due to the fact that proton exchange membrane fuel cell system application is restricted by its limited reliability and durability, an efficient technique for accurate and fast identification of PEM fuel cell faults is highly required. In this study, a PEM fuel cell fault diagnostic method is proposed based on fuzzy logic model and electrochemical impedance spectroscopy. The fuzzy rules used in fuzzy logic model for inferring diagnosis are derived directly from PEM fuel cell faulty mechanisms, thus avoiding inconsistent and even conflicting fuzzy rules used in previous studies. Moreover, its effectiveness in identifying different PEM fuel cell faults and various levels of faults are investigated. Results demonstrate that PEM fuel cell flooding and dehydration can be diagnosed accurately and PEM fuel cell faults at different levels can also be discriminated.