On a fault detection system based on neuro-fuzzy fusion method

A fault detection system of a power plant by means of neuro-fuzzy fusion method is discussed in this paper. Stator temperature together with temperature variations of refrigerant at the entrance and exit of hydrogenerator group are firstly monitored by temperature sensors. The fuzzy system of the aforementioned variables and their membership functions, next, are designed according to expertise knowledge base. Finally, a neuron-fuzzy fusion model is generated by merge the fuzzy system into a neural networks fusion model, and it is proved to be efficient with a correctness ratio of 91% by testing experiments on around one third of overall samples.