基于多元经验模态分解互近似熵及GG聚类的轴承故障诊断

提出了一种基于多元经验模态分解（Multi-EMD）、互近似熵和GG聚类的滚动故障轴承诊断方法。首先，将振动信号进行多元经验模态分解，得到若干个内禀模态函数（IMF）分量和一个趋势项。然后，将IMF分量分别与原始信号进行相关性分析，筛选出前7个含主要特征信息的IMF分量，并将筛选的IMF分量的互近似熵作为特征向量。最后，将特征向量输入到GG模糊分类器中进行聚类识别。通过聚类三维图，对两种算法机械运行的4种状态进行了对比，验证了多元经验模态分解方法不仅可解决采样的不均衡问题，而且可解决EMD算法聚类的混叠问题。

Bearing Fault Diagnosis Based on multi-EMD,cApEn and GG Clustering Algorithm

A new method for rolling bearing fault diagnosis was introduced based on the multi-EMD, cApEn and GG clustering algorithm. The rolling bearing vibration signals were decomposed first by multi-EMD to obtain several intrinsic mode function (IMF) components and a tendency item. Then the first seven IMF components involving the primary feature informations were chosen by the criteria of correlation with the original signals, and the cApEn entropies of each IMF component were composed eigenvectors. Finally, the constructed eigenvectors were put into GG classifier to recognize different fault types. The four kinds of operating states of the machine were presented by means of clustering three-dimensional graph, which instates that the unproportional sampling may be solved by the multi-EMD method and the cluster aliasing of EMD can be further solved.