Real-time fault diagnostic system for a steam turbine generator set by using a fuzzy cerebellar model articulation controller

Abstract

Because a serious fault would result in a reduced amount of electricity supply in a power plant, the real-time fault diagnosis system is extremely important for a steam turbine generator set. A novel real-time intelligent fault diagnosis system is proposed by using a fuzzy cerebellar model articulation controller (CMAC) neural network to detect and identify the faults and failures of critical components. A framework of the fault diagnosis system is described. The model of a novel fault diagnosis system by using a fuzzy CMAC is built and analysed in detail step by step. A case of the diagnosis including three faults is simulated with a fuzzy CMAC. The results show that the real-time fault diagnostic system is of high accuracy, quick convergence, and high noise rejection. Moreover, the model is verified by two examples. It is found that this model is feasible. Finally, the effects of the generalization parameter and address number in fault diagnosis are discussed.