Model based fault detection and isolation of PEM fuel cell

This paper presents a procedure of fault detection and isolation of proton exchange membrane (PEM) fuel cell based on its mathematical model by nonlinear analytical redundancy approach. The model is simplified into five nonlinear state equations representation. The transient phenomena captured in the model include the compressor dynamics, the flow characteristics, mass and energy conservation and manifold fluidic mechanics. Analytical redundancy relations (ARR) are generated and numerical simulation results are given. Two actuator faults and two sensor faults are detected and isolated. The diagnosability of fuel cell system is also obtained.