旋转机械振动信号基于EMD的希尔伯特变换和小波变换时频分析比较

基于经验模态分解(Empirical Mode Decomposition，EMD)的希尔伯特变换(Hilbert Ttransformation，HT)，是先把一列时间序列数据通过经验模态分解，然后经过希尔伯特变换获得频谱的信号处理新方法。介绍了该方法的理论和算法。对仿真和旋转机械油膜涡动故障振动信号分别用基于EMD的HT和基于小波变换(Wavelet Transformation，WT)的时频分析在时域、时频域和频域进行了比较研究，研究结果说明，旋转机械振动信号基于EMD的HT时频分析方法比基于WT的有效。

THE COMPARISON OF VIBRATION SIGNALS' TIME-FREQUENCY ANALYSIS BETWEEN EMD-BASED HT AND WT METHOD IN ROTATING MACHINERY

The method of empirical mode decomposition (EMD) based Hilbert Transformation(HT) is that, first, separate the series data to components with different time scale, say, intrinsic mode function (IMF), using EMD, then apply the Hilbert transformation to every IMFs. The result of the method is the spectrum of the method. The method and algorithm was explained in detail. Emulation and oil film whirling vibration signals were analyzed using the method. The decomposed signals profiles were revealed by IMFs figure. After applied Hilbert transformation to the signals IMFs, the time-frequency and marginal spectrums were obtained. The Wavelet Transformation (WT) method was applied to the signals also. The components of the signals were gained by using wavelet decomposition method. After applied continue wavelet transformation to the signal, the WT-based time-frequency and marginal spectrums were acquired. Compared the two time-frequency method in the domain of time, time-frequency and frequency show that, The EMD-based HT time-frequency analysis of the vibration signal is more sharp and clear than the WT-baseds ,and the WT-baseds frequencies were confused somewhat. The conclusion was deduced that the EMD-based HT time-frequency analysis method is better than the WT-baseds in the field of vibration signals processing in rotating machinery.