Local maximum multi-synchrosqueezing transform for the analysis of time-varying signals

Abstract

Abstract Synchrosqueezing transform (SST) and its improved algorithm are used to process time-varying signals and have been widely developed in the field of signal processing recently. However, processing strong time-varying signals is still a tricky problem. Multisynchrosqueezing transform (MSST) is an excellent time-frequency (TF) analysis technique for processing strong time-varying signals. However, some TF points will not be rearranged using this method. So we propose a new algorithm named local maximum multi-synchrosqueezing transform. The method detects the local maximum of MSST to address the problem existing in MSST. In this way, a near-perfect TF analysis method used to analyze time-varying signals is generated. Simulated signal and experimental data demonstrate the effectiveness of our method.