Time-frequency analysis of I.C. engine vibration signals based on EWT-PWVD

摘要

In order to effectively solve the problem of the contradiction between the cross-interference term and the time-frequency aggregation in the analysis of multi-component signals by Wigner-Ville distribution (WVD), an new time-frequency analysis method based on empirical wavelet transform (EWT) and Pseudo-Wigner distribution (PWVD), named EWT-PWVD is proposed. Firstly, the multi-component signals are decomposed by EWT and a set of intrinsic mode functions (IMF) are obtained. Then each IMF is analyzed by PWVD. Finally, the analysis results of each IMF are linear superposed to reconstruct the time-frequency distribution of the original signal. The simulation results show this method can effectively suppress the cross-interference terms of WVD and maintain its original characteristics. Meanwhile this method is applied to time-frequency analysis of Internal Combustion Engine (I.C. Engine) vibration signals. The result shows this method can clearly depict the representative characteristic of signals and eliminate the cross-interference terms. EWT-PWVD is an effective time-frequency analysis method of signals.