Design Of Wireless Vibration Monitoring And Fault Diagnosis System For Hydrogenerator Unit

During the long-term operation of the hydro-generator unit,it will gradually be affected by various external influences,such as mechanical wear,sediment erosion,cavitation wear and so on.Over time,these electrical or mechanical damages will eventually lead to a reduction in the efficiency of power generation equipment and even accidents,and most of these damages can be reflected by the vibrations generated by the unit's forces.In the collection of vibration data for hydro-generating units,if the traditional wire transmission mode is used,there is a risk that the wire will be immersed in water or cut by the guide vanes,and the way of wireless transmission can get rid of the problem of the line.The complex nonlinear system of water turbine generator sets may be different in different operating conditions of the same fault.The traditional method of fault diagnosis for water turbine generator set based on signal processing only depending on the time-frequency characteristics of vibration signals to identify fault patterns,it will ignore the influence of hydroelectric generating units on the time-frequency characteristics of vibration signals..This project has designed the wireless vibration monitoring system of the turbine generator set,and built the vibration fault model of the turbine generator set,which can transmit the vibration data and quickly diagnose and warn the fault.First of all,the nodes of the wireless vibration monitoring system of the hydroelectric generating unit are mainly divided into wireless vibration sensor acquisition nodes and wireless vibration gathering nodes.The hardware and software of the monitoring system nodes is designed,including hardware selection,circuit design,RTU software design and so on.Complete the wireless vibration monitoring of hydroelectric generating set,and achieve the function of vibration data acquisition and wireless transmission.Secondly,feature extraction and feature selection for vibration characteristic parameters of hydro generator set can not only improve the efficiency of fault diagnosis,but also improve the accuracy.In this paper,the fast independent component analysis(FastICA)algorithm can be used to extract the characteristics of the non-stationary weak vibration signal,and the large interference signal and the strong noise can be removed.The Relief algorithm in the feature selection filter and the maximum minimum redundancy algorithm are used to reduce the feature parameters,and the support vector machine(SVM)SVR is used to identify the feature selection.The method proposed in this paper has higher diagnostic accuracy,better timeliness and is more suitable for feature extraction and selection of hydroelectric generating set.Finally,the nonlinear state estimation(NSET)is used as a nonlinear and nonparametric multi-dimensional modeling diagnosis method,which is used to model the vibration data of the turbine generator set.In this paper,Matlab simulation software is used to complete NSET modeling and fault diagnosis.Finally,the diagnosis results are analyzed to verify the feasibility of NSET model.