Order Spectrum Analysis for Bearing Fault Detection via Joint Application of Synchrosqueezing Transform and Multiscale Chirplet Path Pursuit

Abstract

Order tracking has become one of the most effective methods for fault detection of rotating machinery under the time-varying shaft speed conditions. The transient phase estimation is very important for order tracking, especially when the tachometer installation is not convenient. The transient phase is usually obtained by integrating the instantaneous frequency (IF), so the IF estimation has attracted a great deal of concerns. This article describes a new IF estimation method based on the joint application of the synchrosqueezing transform (SST) and the multiscale chirplet path pursuit (MSCPP) method. The SST method as its high frequency resolution merits is used to estimate the frequency parameters for the parameter settings of the MSCPP method, that will resolve the high computation problem of the MSCPP method to a certain degree, so as to extensively use the high accuracy of the MSCPP method in IF estimation. The order spectrum based on the estimated IF can provide the demodulation information for the bearing fault diagnosis. The performance of the proposed method has been validated by both simulation and experimental data.