Damping identification using synchrosqueezed wavelet transform

Synchrosqueezing is a procedure for improving the frequency localization of a continuous wavelet transform (CWT). This research focuses on using a synchrosqueezed wavelet transform (SWT) to determine the damping ratios of a vibrating system using a free-response signal. While synchrosqueezing is advantageous due to its localization in the frequency domain, damping identification with the SWT is not sufficiently accurate. Although the SWT builds on an existing CWT, which is renowned for its robustness even in case of very noisy signals, the SWT performs notably worse when noise is embedded in signal. This work focuses on describing the origin of SWTs worse performance in case of noisy signals. It demonstrates that the uncertainty arises with computation of preliminary frequencies and it is exposed as a result of the reallocation criteria in SWT. To improve the SWT, a modified synchrosqueezing criterion, the average SWT, is introduced and tested against the noise in the signals.