基于EEMD散布熵-LPP的滚动轴承故障诊断方法  

Fault diagnosis method of rolling bearings based on EEMD spread entropy and LPP

作　　者：蒋永华[1,2] 张晓迪 黄涛涛 焦卫东 李刚[1] 高昭[1] 施继忠[1] JIANG Yonghua;ZHANG Xiaodi;HUANG Taotao;JIAO Weidong;LI Gang;GAO Zhao;SHI Jizhong(Key Laboratory of Urban Rail Transit Intelligent Operation and Maintenance Technology and Equipment of Zhejiang Province,Zhejiang Normal University,Jinhua 321004,China;Xingzhi College,Zhejiang Normal University,Lanxi 321100,China)

机构地区：[1]浙江师范大学浙江省城市轨道交通智能运维技术与装备重点实验室,浙江金华321004 [2]浙江师范大学行知学院,浙江兰溪321100

出　　处：《浙江师范大学学报（自然科学版）》2020年第4期381-387,共7页Journal of Zhejiang Normal University:Natural Sciences

基　　金：国家自然科学基金资助项目(51405449,51575497);浙江省城市轨道交通智能运维技术与装备重点实验室自主研究课题(ZSDRTZZ2020002)。

摘　　要：针对滚动轴承故障分类中采用单一尺度熵值难以完全表征故障特征的问题,基于散布熵(DE)、集合经验模态分解(EEMD)、局部保留投影算法(LPP)、k-近邻分类算法(KNN),提出了一种基于EEMD散布熵-LPP的滚动轴承故障诊断方法.首先对信号进行EEMD分解提取其前9个IMF分量作为特征信号,分别计算其散布熵值作为特征参数构建高维特征集,接着利用LPP进行降维获得低维特征集,最后输入KNN分类器进行分类识别.对实际滚动轴承的正常、内圈故障、外圈故障、滚动体故障样本进行了分类识别,并与其他几种方法进行对比,结果表明:该方法具有更高的分类准确率.For the fact that a single-scale entropy value was difficult to fully express the characteristics of rolling bearing failure,a rolling bearing fault diagnosis method was proposed based on ensemble empirical mode decomposition(EEMD),dispersion entropy(DE)and local reservation projection(LPP).The method adopted dispersion entropy as the characteristic parameter,the first nine intrinsic mode functions(IMFs)of the signal were extracted with EEMD and used as the characteristic signals;their dispersion entropies were calculated and used as the characteristic vectors and the sample set was obtained.Then,LPP was used to extract the initially obtained feature set,and finally the k-nearest neighbor(KNN)was used to identify the characteristic extraction method.The results showed that the method had higher classification accuracy by using the rolling bearing fault sample for test verification and compared with several other methods.

关 键 词：散布熵 集合经验模态分解 流形学习 故障诊断 滚动轴承 

分 类 号：TH165.3[机械工程—机械制造及自动化] TN911.2[电子电信—通信与信息系统]