Bearing Condition Monitoring of Induction Motor Based on Discrete Wavelet Transform & K-nearest Neighbor

Abstract

A novel technique for the early detection and diagnosis of bearing faults in three phase induction motor (IM), based on discrete wavelet transform (DWT) and k nearest neighbors (KNN) is proposed in this paper. The Three-phase IM are the back bone of most industrial processes. The bearing faults are amongst the most probable mechanical faults that affect the operation of IM. The early detection of bearing condition prevents the possible loss in terms of revenue as well as the downtimes of the industrial process. The proposed method is validated with real voltage and current signals acquired using laboratory experimental setup for a healthy and two unhealthy bearings. The results exhibit good accuracy and ability to differentiate between healthy and unhealthy bearing conditions.