A tacholess order tracking method based on inverse short-time Fourier transform and singular value decomposition

The ability of order tracking on improving the performance of bearing fault diagnosis has been widely recognized. The key work of tacholess order tracking (TOT) is to extract a component from the raw signal that is closely connected to the shaft rotating and then obtain the phase from the component. In this paper, a TOT method is proposed based on inverse short-time Fourier transform (ISTFT) and singular value decomposition (SVD). The component of the original signal shows strong periodicity of the shaft rotation frequency is selected and roughly filtered in the time-frequency domain after the STFT. Then, the ISTFT is applied to reconstruct the selected component into the time domain. Next, SVD is applied to refine the roughly filtered component. Finally, the unwrapped phase of the extracted component is used to carry out the resampling operation. The performance of the method is tested using vibration signals collected from a test rig of a high-speed vehicle traction system