22nd IEEE Signal Processing and Communications Applications Conference (SIU), Trabzon, Türkiye, 23 - 25 Nisan 2014, ss.991-994
In this study, artificial defects in various diameters are formed on inner race, outer race and ball bearing which are essential components of a bearing and vibration signals are collected by a data acquisition card from bearing-shaft setup. The signals acquired are decomposed from noise with wavelet transform; thus vibration signal resulting from normal operation of the system is obtained The energy of noisy and noise-free signal is calculated and the wavelet coefficients that will be used in classifying are obtained. As a conclusion of experimental studies, the technique based on wavelet transform coefficients accomplishes the classifications of different bearings fault types successfully.