Fault Diagnosis and Accommodation System with a Hybrid Model for Fuel Cell Power Plant

As a part of U.S. distributed generation program, the concept of hybrid fuel cell power plants has shown its application potential and is already under commercialization. In the hybrid direct fuel cell and turbine (DFC/T) plant, the control system is an essential component that guarantees reliable and efficient operations. Being an effective complement for the local control scheme, a fault diagnosis and accommodation system based on fuzzy logic is implemented with the help of neural network augmenters, which improve the accuracy of the nominal model of the fuel cell plant. As a quantitative approach, the fuzzy fault diagnosis and accommodation system has considerably lower realization complexities than an analytical and detailed fault regulator. The system structures and design methods are discussed in this paper. Model augmentation and simulation results are presented to verify the performance of the overall system.