Fault diagnosis based on D-S evidence theory in the application of power plant

The monitoring of operation states and fault diagnosis system of turbines in power plant are significant to guarantee the units long-term safety and economic operation. The fault diagnosis of turbines is influenced by various factors. In order to enhance the accuracy and reliability of fault diagnosis, it needs to utilize the different information from various sensors. This paper has presented a fault diagnosis system based on D-S evidence theory. Firstly, the different data transformed through fuzzy membership function are as the inputs of neural network. The outputs of neural network are the primary fault diagnosis, which usually can determine the type of faults. But in some cases, it is unable to determine the fault type accurately. Therefore the information fusion is applied to accomplish the further fault diagnosis. With D-S evidence theory, all possible kinds of information can be used to improve the accuracy of diagnosis. This method has been successfully applied in the fault diagnosis of condenser. Compared with the general method of FNN, this approach can enhance the accuracy of fault diagnosis, especially for reducing the uncertainty in the fault diagnosis.