Bearing Fault Classification Based on Envelope Analysis and Artificial Neural Network

Bearings are among the most stressed components of industrial machines and represent a frequent source of failure. The diagnosis of these defaults is very important in the domain of predictive maintenance. In this paper, kurtogram technique is applied as an alternative method to determine the optimal band-pass filter characteristics used for envelope analysis. The signal envelope obtained by Hilbert transform and the FFT transform allow the features extraction. Then, an artificial neural network has been used to classify fault type of the rolling element.