PEM Fuel Cell System Robust LPV model-based Fault Diagnosis

In this paper, the problem of robust fault diagnosis in a PEM Fuel Cell (PEM FC) system is addressed. Fault detection is based on a LPV interval observer able to face the variations of parameters with the operating while preserving the linear structure of the model. At the same time, this approach addresses the robustness problem against modeling uncertainty included as unknown parameters whose value is bounded by intervals. Parametric uncertainty effects are propagated to the observer output using a zonotope representation to approximate and propagate the set of possible states. Fault isolation is based ona set of structured residuals that are analysed using a relative fault sensitivity analysis approach. The proposed methodology of fault diagnosis is applied to a PEM Fuel Cell system.