The actual measured vibration signal of the defective mechanical equipment is generally non-stationary. Extracting effective and useful features from non-stationary signals has always been an important research topic in fault diagnosis. A novel time–frequency analysis method, called time-reassigned synchrosqueezing extracting S-transform has been proposed in this study. Compared with traditional time-frequency analysis methods, the method can not only extract the useful information of non-stationary signals well, and obtain a more concentrated time-frequency representation, but also has better noise robustness and lower time consumption. The analysis of numerical signals verified the great performance of this method. Finally, time-reassigned synchrosqueezing extracting S-transform (TSSEST) is applied to perform the spectral decomposition of a bearing fault data. The numerical and experimental signals are used to show the effectiveness of our method.

S-Transform; time–frequency analysis; Fault Detection;