A time-frequency analysis of non-stationary signals using variation mode decomposition and synchrosqueezing techniques

• Published in: 2019 Prognostics and System Health Management Conference (PHM-Qingdao) pp. 1 - 6
• Main Authors: Ni, Chao, Lin, Tian Ran, Peng Xing, Jin, Xue, Jun Zhou
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.10.2019
• Subjects: Entropy; Fault diagnosis; Fourier transforms; kurtosis; Rényi entropy; Signal reconstruction; Signal resolution; Synchroqueezing transform; Time-frequency analysis; time-frequency reconstruction; Variational mode decomposition
• DOI: 10.1109/PHM-Qingdao46334.2019.8943036

Time-frequency analysis (TFA) technique is an effective method in the analysis of non-stationary signals with multiple signal components. Nevertheless, the current TFA methods such as short-time Fourier transform (STFT) and synchrosqueezing transform (SST) suffer from low signal resolution and diffused energy problems in the processing of multi-component signals. To resolve such shortcoming, a new signal analysis technique utilizing a combined variational mode decomposition (VMD) and synchrosqueezing transform (SST) is proposed in this study. VMD is used first to decompose the signal into several intrinsic mode functions (IMFs). Kurtosis values of the IMFs are calculated to determine the fault sensitive components to be used in the signal reconstruction. The noise reduced reconstructed signal is then analyzed using SST to produce a clearer and energy concentrated time-frequency representation (TFR) for an improved bearing fault diagnosis. The validity of the presented technique is verified utilizing a computer simulated signal and a bearing defect signal acquired from the experiment.