Time-frequency analysis method of bearing fault diagnosis based on the generalized S transformation

The generalized S transform (GST) can flexibly adjust the change trend of the fundamental window function according to the frequency distribution characteristics and the time-frequency emphasis of vibration signal. Also, this transform can accelerate or slow down the time-band width change along with the frequency, in order to make the amplitude of the fundamental window function present the various nonlinear changes. These features are very constructive significance for signal analysis and processing. A time-frequency analysis method based on the generalized S transformation is introduced to the fault diagnosis of bearing. And the application principle and the steps of method applied in fault diagnosis are given. Simulated signals and the actual bearing fault signals from inner race, outer race and rolling body are processed to verify the validity of the proposed method. Results show that the method can effectively enhance the resolution of vibration signal not only in time domain but also in frequency domain. The fault characteristic frequency can be extracted from the reconstructed signal, and the status of bearing and the fault type of bearing can be obviously distinguished. The presented time-frequency method effectively improves the accuracy of the fault diagnosis of bearings.