Vibration Analysis and Fault Diagnosis of Induction Motor Bearing Using Artificial Neural Network (ANN)

Induction motor bearing faults are one of the main causes of catastrophic failure of machines. Thus, detection and diagnosis of faults in bearings is very crucial for the reliable. This paper focuses on fault diagnosis of induction motor bearing having localized defects using Daubechies wavelets-based feature extraction. In present study Machinery Fault Simulator (MFS) test rig used for fault diagnosis of NSK-6203 deep groove ball bearing. Vibration signals collected from the various bearing conditions- healthy bearing (HB), outer race defect (ORD), inner race defect (IRD), ball defect (BD) and combined bearing defect (CBD). The extraction of statistical features carried out using various Daubechies wavelet coefficients from raw vibration signals. Lastly, the bearing faults are classified using these statistical features as input to Artificial Neural Network (ANN) technique used for faults classifications. The test result shows that ANN identifies the fault categories of rolling element bearing more accurately for Db4 and has a better diagnosis performance as compared to other Daubechies wavelets with ANN classifier.