Rotor Fault Diagnosis Research Of High Voltage Asynchronous Motor Based On Intelligent Method

With modern science technology developing and the scale of power system increasing, motors play a more and more important role in the industrial production. The motor fault will not only damage itself, affect the normal operation of the whole system.but only cause serious economical losses and significant social impact,even endanger personal safety in the serious case. The motor fault diagnosis is of great important and significant practical value in engineer,By researching on monitoring fault diagnosis of motor,discovering and eliminating the fault in time.It is of great significance to improve the production efficiency and ensure the reliable operation of motor.On the basis of discuss the motor fault diagnosis domestic and overseas,It carried on a simple analysis on the basic structure of rotor of high voltage asynchronous motor and rotor fault. Thesis proposed two motor fault diagnosis method which based on different mechanism, "plain-type bayesian network topology model" and "rotor fault diagnosis model of wavelet neutral network". Because rotor fault information of power system exist uncertainty,on the basis of traditional method, by the way integrating sample and prior information, the model based on rotor fault diagnosis system which correspond to the fault symptoms develop.Bayesian network is one of the most effective module in current expressing and reasoning field, It confined this method can effectively handle information under the uncertainty of motor faults through practicing handing to motor. By the same time, The fault signal contain a large numbers of short--term and sudden qualitative composition. Traditional Fourier transform in frequency domain just analysis the amplitude and density of the frequency cannot detect singular point of time-domain signal. The result cannot reach the requirement of fault signal extraction. Wavelet neutral network develop the fault diagnosis module based on motor vibration signal, using the wavelet time-frequency technology filter the no-noising and extract the fault feature,then the BP network finish fault recognition to achieve the goal of fault diagnosis.Finally,for the typical fault of asynchronous motor:rotor imbalance,quality bearing friction and rotor is in false direction,using the above two different fault diagnosis method to develop module.It proves the two different method mentioned can effectively diagnosis high voltage asynchronous motor fault and improve the precision of motor fault diagnosis.