A parameterized iterative synchrosqueezing transform for the analysis of noise contaminated non-stationary signals

Synchrosqueezing transform (SST) is a powerful post-processing tool in the analysis of non-stationary signals, though it has limited capacity in the handling of strong time-varying signals or signals with low signal-to-noise ratio (SNR). In this study, a parameterized iterative SST is proposed for an accurate machine fault diagnosis. Firstly, a parameterized chirplet transform is used to capture the fast-varying instantaneous frequencies (IF) of a strong time varying signal. Secondly, the TF coefficients are reassigned to the IF trajectory by an iterative SST technique. Finally, the blurry transition between two neighboring TF points is eliminated by the construction of a local maxima operator. The effectiveness of the current technique is examined using two simulated non-stationary signals and two sets of machine defect data acquired under varying speed condition. The results show that the current technique can produce a highly energy concentrated time frequency result for an accurate mechanical fault diagnosis under either strong speed variation or/and noise contamination conditions. The effectiveness of the technique is also validated by comparing the result with those obtained via other TFA techniques. The result showed that the current technique can render the most accurate result within the TFA techniques used in this study.