Model-based humidity observer for vehicle proton exchange membrane fuel cell based on a new sliding mode observer estimation technique

The internal water content of the proton exchange membrane fuel cell (PEMFC) stack significantly affects its performance and duration. To achieve optimal humidity level, real-time monitoring of the humidity of the cathode and the anode is crucial for the PEMFC. As far as we know, the relative humidity of the PEMFC can be estimated via an observer based on the measurable output voltage. However, the rapid changes in vehicle operating conditions pose a challenge to the estimation accuracy. Different from the existing sliding mode observers (SMOs), this study pays more attention to the testing noise of voltage measuring sensor. To reduce the noise influence and improve the estimation accuracy, a new sliding mode observer (NSMO) combined with an online Kalman filter is presented based on the dimensionless model of the PEMFC in this study. The applicability of the proposed NSMO in estimating the cathode relative humidity and the anode relative humidity is illustrated by the simulation results. At the same time, comparisons between the NSMO and the online Kalman filter are provided which show a substantially higher estimation accuracy and lower cost for the NSMO. Furthermore, the final results indicate the mean absolute deviations of the cathode and the anode relative humidity for the NSMO are, respectively, reduced to 0.39% and 0.28%.