EMD中有效IMF选取方法的研究

Methods of selecting valid IMF in EMD

阐述了经验模态分解(EMD)中有效本征模态分量(IMF)的3种选取方法(相关系数法、KL散度法以及能量守恒与频率散度结合法),并将该3种方法应用于某台汽轮机典型动静碰磨和油膜振荡信号的分解试验中。分析试验结果表明,将针对低频的能量守恒法与针对高频成分的频率散度法相结合的选取方法在区分度方面优于相关系数法,在适应性方面优于KL散度法,总体的判断稳定性均优于相关系数法和KL散度法。

Three methods （correlation coefficient method, KL divergence method, energy conserva- tion combined with frequency divergence method） for extracting valid intrinsic mode function （IMF） from empirical mode decomposition （EMD） were presented. Moreover,decomposition test was performed on typical rubbing （between dynamic and static parts） and oil film oscillation sig- nal of a steam turbine, by using the above three methods. The results show that, the selection method which combing the energy conservation method for low frequency component with the frequency divergence method for high frequency component was superior to the correlation coeffi- cient method in discrimination, and to the KL divergence method in adaptability,and to both of them in overall stability of judgements.