FAULT ANALYSIS IN SELF ALIGNING BALL BEARING BY WAVELET TRANSFORM BASED FEATURE EXTRACTION USING NEURAL NETWORKS

The objective of the present work is to classify different individual defects in case of self aligning ball bearing by using statistical tools coupled with a machine learning technique. The analysis of the generated results is then made and a better understanding of theoretical observations has been put in. To study fault in the bearing, vibration analysis procedure has been undertaken. In the present study, inner race and rough surface defects have been considered. The useful features of the vibration signal have been extracted by using Wavelet Transform which are then used as input to Neural Network algorithm for classification.