Fault Diagnosis of a Single Ball Bearing System Based on Vibration Analysis Using Neural Network

Improving Fault Diagnosis of Single Ball Bearing Systems through Vibration Analysis and Neural Networks

Abstract

The defects in the induction engine bearing are a significant cause of disastrous machinery failure. Detection and detection of flaws in coats is also very critical for safety. This paper focuses on the malfunction evaluation through the usage of wavelet extraction induction engine carrying localized defects. This research uses the research rig for the defective diagnostic of deep ball bearing NSK-6203 for the machinery fault simulator (MFS) examination. Vibration signals obtained from the different conditions of carrying-healthy bearing (HB), outer race defect (ORD), inner race defect (IRD). The elimination of mathematical characteristics from raw vibration metrics utilizing separate wavelet Daubechies coefficients. Finally, these mathematical characteristics are listed as an input to the technique for classification of defects in the artificial neural network (ANN). The outcome of the test indicates that ANN defines more correctly the default groups of rolling components for Db4 and has a higher diagnostic efficiency than for other Daubechy-classified waves.