A Novel Approach for Detection of Bearing Defects from Noisy Vibration Signal

Abstract

Vibration signal of a defective bearing carries fault related information. The aim of this paper is to develop a signal processing methodology that identifies the presence of defect from bearing vibration signal subjected high background noise. A simulated vibration signal considering inner race defect in a deep groove ball bearing with low signal to noise ratio has been generated and investigated. A technique involving CWT of vibration signal and post-processing though FFT has been adopted to analyze the signal. Results show that proposed methodology can yield the presence of inner race defect prominently from a noisy vibration signal.