ANN based Evaluation of Performance of Wavelet Transform for Condition Monitoring of Rolling Element Bearing

Bearings are one of the critical components in rotating machines and the majority of the failure arises from the defective bearings. Bearing failure leads to failure of a machine and unpredicted productivity loss for production facilities. Hence, bearing fault detection and diagnosis is an integral part of the preventive maintenance procedures. In this paper vibration signals for three conditions of a deep groove ball bearing Normal (N), defect on inner race (IR) and defect on outer race (OR) were acquired from a customized bearing test rig, under one load and two speed conditions. Discrete Wavelet Transform (DWT) has been used for vibration signal analysis. The statistical features extracted from the dominant wavelet coefficients are used as inputs to ANN classifier to evaluate its performance. The vibration signals have also been denoised using a new thresholding scheme. A comparison of ANN performance is made based on raw vibration data and denoised data. The ANN performance has been found to be comparatively higher when denoised signals were used as inputs to the classifier. Also various mother wavelet functions (Db8, Db4, Db44 and Sym10) were used to analyze the denoised vibration signals and their performance has been evaluated using the ANN classifiers.