Detection of induction motor bearing damage with starting current analysis using wavelet discrete transform and artificial neural network

Bearing damage in induction motor is one of the most common fault. The type of bearing damage itself consists of damage to the inner-race, outer-race and ball bearing. The occurrence of this bearing damage may cause increased vibration, temperature rise and may cause damage to the shafts, rotor and stator. To speed up the repair process, bearing damage detection should be classified according to the type of damage occurring. In this study, bearing damage will be detected by transient current analysis using discrete wavelet transform method. To determine the occurrence of damage, processing of transient current signals using discrete wavelet transforms performed by comparing the signal sub-band frequency at normal bearings and during fault. Furthermore, artificial neural networks are used to provide information on classification of types of fault. Analysis of result show that the presentage of successness classification as 100% for inner-race damage, 98% for outter-race damage and 100% for ball bearing damage. With the classification of damage to this bearing, it is expected to simplify and speed up the repair process.