Adaptive Condition Monitoring for Fuel Cells Based on Fast EIS and Two-Frequency Impedance Measurements

This article proposes an adaptive condition monitoring method for proton exchange membrane fuel cells based on fast electrochemical impedance spectroscopy and two-frequency impedance measurements. First, an impedance measurement system is developed to achieve fast electrochemical impedance spectroscopy and impedance measurements of single frequency. Second, two characteristic frequencies of the fuel cell stack are adaptively extracted from the impedance spectrum. With the two characteristic frequencies, an online state classification algorithm is proposed based on a multiclass linear discriminant classifier to realize the condition monitoring for fuel cells. Finally, the results are validated experimentally on a 3-kW and a 400-W fuel cell stack to verify the effectiveness, rapidity, and migrability of the proposed method.