Fault detection and isolation for PEMFC systems under closed-loop control

In this work, a model-based fault detection and isolation (FDI) is developed for proton exchange membrane (PEM) fuel cell (FC) stack that is under feed-forward plus feedback control. The fault detection is achieved using an independent radial basis function (RBF) network model, whilst the fault isolation is based on the RBF classification. The novelty is that the RBF model of independent mode is used to predict the future outputs of the FC stack and a RBF classifier is used to classify five types of fault introduced to the PEMFC systems. To validate the method, a benchmark model developed by Michigan University is used in the simulation to analyze the effectiveness of the method for actuator, component and three sensor faults. The FDI results corresponding to those scenarios show that the simulated different types of fault are successfully detected and isolated.