Research On Online Monitoring And Fault Diagnosis Of Generator Set

In order to meet the requirement of a online monitoring project for turbine-generator from Thermoelectric Division Of Baling Petrochemical Co., Ltd , a online monitoring and fault diagnosis system consists of a embedded data acquisition module ,a monitoring subsystem in workstation and Web service fault analysis software is presented. The system is designed by means of advanced signal processing methods, modern electronic measuring & control technology and popular network structure.The article begin with a brief summary of the contents and applications of condition monitoring technology, which is followed by the analysis of the status and development trend of domestic and foreign industry. Then we discuss the latest relevant technologies. Through several aspects of this discussion, the system used Client-Server and Browser-Server software framework and the features and specifications is introduced.This paper mainly discussed the algorithm of fault diagnosis. The fault diagnosis realized in the system can be classified into three groups: time domain analysis, frequency domain analysis and time-frequency analysis. Both stationary and non-stationary signals are suitable to be analyzed by the system. The running state of generator set can be real-time observed in both time domain and domain. Take the advantage of the time-frequency analysis methods such as Wigner-Ville Distribution, This system is characterized by its ability to analysis the exact time and position of an actual fault. The article explores the theory and discusses the physical meaning of the wavelet. Than the excellent characteristics of its appliance in the fault diagnosis is deeply discussed. The decomposition singularity signal and threshold de-noise are realized in the system.An expert system based on L-M BP neural networks is studied. Energy eigenvector from vibration signals which have been decomposed by tri level wavelet are used as the input of expert system. Stimulation results demonstrate the convergences of expert system are well, and it can avoid falling into local minimum. It also has better fault tolerance and high stability.The online monitoring and fault diagnosis of generator set system has successfully applied in Thermoelectric Division Of Baling Petrochemical Co., Ltd, the practice show it is stable and reliable. The actual operate is show in the end of the paper.