The Fault Diagnosis Of Motor Bearing

Motor is the most important equipment of mine, and its reliable operation is directlyrelated with the refinery’s safety and economic efficiency. The reliability of motor canbe improved by monitoring and fault diagnosis. So it is significant to design a reliablemonitoring and fault diagnosis system.Rolling bearing is one of the most important components of mine motor. Therelevant statistics shows that30%of the motor faults are caused by the rolling bearing.In this paper, it is main to monitor the rolling bearing vibration condition to reach thepurpose of fault diagnosis. So the mechanism of rolling bearing fault vibration andsome diagnostic basis which includes the form of rolling bearing fault, fault reasons,common diagnostic methods are studied in the thesis. The hardware was designed bythe data collection card PCI-1710HG produced by Yanhua company, combining withthe accelerometer LC0159and the signal disposal. The corresponding data collectionsoftware and fault detection software were designed through the virtual instrumentsLabVIEW.An ensemble approach based on wavelet packet entropy and clustering analysis ispresented to diagnose faults in the rolling bearing vibration signal research. In order tocertify the validity of the proposed method, the wavelet packet entropy and clusteringanalysis approach were applied in bearing fault diagnosis experiment of West ReserveUniversity. The experimental results show that the recognition rate of the proposedapproach is much higher than the K-means in rolling bearing fault recognition.Combined with the characteristics of fault diagnosis, a remote Rolling bearing faultdiagnosis system was designed by the virtual instrument software LabVIEW andMATLAB toolbox.