Bearing fault diagnosis using Gaussian mixture models (GMMs)

This paper presents a novel method for bearing fault diagnosis based on wavelet transform and Gaussian mixture models (GMMs). Vibration signals for normal bearings, bearings with inner race faults, outer race faults and ball faults were acquired from a motor-driven experimental system. The wavelet transform was used to process the vibration signals and to generate feature vectors. GMMs were trained and used as a diagnostic classifier. Experimental results have shown that GMMs can reliably classify different fault conditions and have a better classification performance as compared to the multilayer perceptron neural networks.