Research On Fault Diagnosis Method Of Hydraulic Turbine Governing System Using Artificial Neural Network

Hydraulic turbine is one of the most important equipments in hydropower station, and it influences directly the economic and social benefits of the hydropower station. As one of the key equipments of hydraulic turbines, the hydraulic turbine generator plays a great role on normal production of electric power. Therefore, it's necessary to study the fault diagnosis method of hydraulic turbine governing system both in theory and practice.At present, fault diagnosis method of hydraulic turbine governing system has many disadvantages such as long time need and low accurate ratio. Aiming at the urgent needs of remote monitor system in the hydropower station, the method of D-S evidential reasoning based on neural networks was introduced to study the fault diagnosis of hydraulic turbine governing system. The main contents are as follows:(1) The basic theory and status of the fault diagnosis method is discussed, and the characteristics of hydraulic turbine governing system are studied. The effect and problems of using BP neural network and RBF neural network to detect the faults of the hydraulic turbine are also analyzed in detail.(2) The theory of using D-S evidential reasoning to make data fusion about the fault diagnosis results with different artificial neural networks techniques is mainly set up in the paper, which is applied in the hydraulic turbine governing system. Experimental results show that the approach is effective and efficient.(3) The VC++ and SQL Server are used to program the software, and the proposed approach is proved successfully by Xianju Beiao hydropower station.