Fault detection and isolation based on feedforward-feedback control for oxygen excess of fuel cell stack

Feedforward and feedback control is developed in this work for proton exchange membrane (PEM) fuel cell (FC) systems to do fault detection and isolation (FDI). The fault detection is achieved using a radial basis function (RBF) network model, whilst the fault isolation is based on the RBF classification. One actuator fault, one component fault and three sensor faults have been introduced to the PEMFC systems experiencing faults of ±10% from the nominal values in the FC stack. To validate the results, a benchmark model developed by Michigan University is used in the simulation to investigate the effect of these five faults. The developed RBF model is tested on MATLAB R2010b/Simulink environment. By using this method, the RBF network able to detect and isolate all five faults accordingly and accurately under feedforward-feedback control strategy.