Design of Sliding Mode Observer for Internal Oxygen State of PEM Fuel Cell

The lifetime and efficiency of proton exchange membrane fuel cells (PEMFCs) strongly depend on internal oxygen state monitoring. To address the imbalance between the observation performance and computational cost of the existing observers, this paper proposed a reduced-order sliding mode observer. Firstly, an observation-oriented reduced-order PEMFC dynamic model is established, and then an adaptive sliding mode observer was constructed on the basis of this model. Subsequently, the full-state reconstruction was carried out by the easy-to-measure variables, and finally the reconstructed variables were used to realize the accurate estimation of the oxygen excess ratio (OER). The simulation was executed and the results demonstrate that the proposed observer reduces the simulation time by 18.46% and 22.06% compared with Kalman Filter observer and original sliding mode observer (SMO), respectively, while ensuring the dynamic response, steady-state error and robustness of the unreduced-order SMO. This research can provide accurate and useful information for oxygen flow control and fault diagnosis.