改进的HHT方法及其在旋转机械故障诊断中的应用

针对希尔伯特-黄变换(Hilbert-Huang Transform,HHT)方法中存在的模态混叠和虚假固有模态函数(Intrinsic Mode Function,IMF)问题,提出一种基于总体包络均值经验模态分解(Ensemble Envelop Mean Empirical Mode Decomposition,EEMEMD)和虚假模态函数剔除算法相结合的改进HHT方法。该方法利用EEMEMD可准确反映加噪后信号的自身变化,一定程度上中和残留在各模态分量间的噪声,获得无模式混淆的较纯净的IMF分量。同时,通过基于归一化能量熵值的虚假模态函数剔除算法可有效剔除噪声干扰成分和迭代误差分量,从而提高信号特征提取的准确性。通过仿真分析和转子不对中故障诊断的工程实例表明,改进HHT方法能够较好地抑制模态混叠问题并有效剔除同故障无相关的虚假IMF,实现对旋转机械故障的有效诊断。

Improved HHT method and its applications in rotating machinery fault diagnosis

In order to solve the problem of Hilbert-Huang transform (HHT), which has mode mixing and false intrinsic mode function (IMF), an improved HHT method combining ensemble envelop mean empirical mode decomposition(EEMEMD) and spurious mode function elimination algorithm was proposed. The method uses EEMEMD to accurately reflect the self-change of the signal after adding noise, neutralizes the residual noise contained in modal components partly, and obtains a purer IMF component without mode mixing. At the same time, the false mode function elimination algorithm based on normalized energy entropy can effectively eliminate noise interference components and iterative error components, so as to improve the accuracy of signal feature extraction. The simulation analysis and engineering examples of rotor misalignment fault diagnosis were compared.The result show that the improved HHT method can suppress the mode mixing problem better and effectively eliminate the false IMF, which is not related to the fault,to achieve effective fault diagnosis of rotating machinery.