中文摘要：

针对传统的局部均值分解（LMD）方法不能有效提取微弱高频信号成分的问题,提出了一种基于微分的微分局部均值分解（DLMD）方法,在此基础上,将DLMD、样本熵和模糊聚类分析相结合,提出了一种基于DLMD样本熵和模糊聚类的滚动轴承故障诊断方法。该方法首先对滚动轴承振动信号进行微分局部均值分解,得到若干具有物理意义的乘积函数（PF）分量,然后求取各PF分量的样本熵并将其作为特征向量,最后通过模糊聚类对特征向量进行识别分类。实验结果表明,基于DLMD样本熵和模糊聚类相结合的方法能够准确、有效地对滚动轴承故障信号进行识别分类。

英文摘要：

In view of the problem that the traditional local mean decomposition（LMD）was difficult to effectively extract the weak high frequency signal components,a method of DLMD was put forward.A new approach for rolling bearing fault diagnosis based on the combination of DLMD,sample entropy and fuzzy clustering was proposed.Firstly,rolling bearing vibration signals were decomposed with DLMD to obtain a certain number of product function（PF）components which had physical meaning.Then the sample entropies of the PF components were calculated and used as the eigenvectors.Finally,the eigenvectors were recognized and classified through the fuzzy clustering.The experimental results show that the method based on the combination of DLMD,sample entropy and fuzzy clustering can be used to recognize and classify rolling bearing fault signals accurately and effectively.