In order to estimate fuel-cell degradation status on-line and inexpensively, a diagnostic technique based on relay feedback is developed. The technique can obtain critical parameters within seconds of start-up and is robust to measurement bias. Electrochemical impedance spectroscopy (EIS) is a popular laboratory technique to perform diagnostics on electrochemical systems such as fuel cells, but its application to real-life fuel-cell systems is difficult because of the size and cost of the apparatus. In this study, we present a more detailed equivalent-circuit model for a PEM fuel cell, able to explain the positive reactance shown at low frequencies. Some of these characteristics, measured at several stages during an Accelerated Stress Test (AST), progress gradually with catalyst degradation, providing an effective prognostic variable. In order to measure these characteristics, a relay-based feedback excitation algorithm is developed to estimate the low-frequency intercept in the Nyquist plane of the cell impedance without resorting to a full-fledged EIS. The simulations indicate that the algorithm converges to an estimate within about 5 seconds, and is robust to bias. The algorithm can be run within the standard control system that fuel cells are usually equipped with, with no additional hardware.

Data published under ODbL.