基于EEMD能量比和GG聚类的滚动轴承故障诊断  

Fault Diagnosis of Rolling Bearing Based on EEMD Energy Ratio and GG Clustering

作　　者：马丽华 朱春梅[1] 赵西伟 MA Li-hua;ZHU Chun-mei;ZHAO Xi-wei(Key Laboratory of Modern Measurement and Control Technology,Beijing Information Science and Technology University , Beijing 100192,China)

机构地区：[1]北京信息科技大学现代测控技术教育部重点实验室,北京100192

出　　处：《组合机床与自动化加工技术》2020年第5期21-26,共6页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：国家自然科学基金资助项目(51275052);北京市自然科学基金重点项目资助(3131002);京津冀自然科学基金基础研究合作项目(J170004)。

摘　　要：针对滚动轴承故障特征提取困难导致故障类型难以辨识的问题,提出基于集合经验模态分解(Ensemble Empirical Mode Decomposition,EEMD)和GG(Gath-Geva,GG)聚类的轴承故障诊断方法。首先,使用EEMD分解方法对轴承的振动信号进行分解,结合相关系数原则提取含有主要故障信息的4个固有模态函数(IMF)分量,计算其能量百分比作为特征值,再用GG聚类对特征值进行聚类分析。通过仿真验证了GG聚类的优越性,然后采用文中提出的GG聚类方法与FCM聚类、GK聚类对轴承故障数据的聚类效果进行对比分析,验证了文中所提方法在滚动轴承故障识别中的可行性。Aiming at the difficulty in extracting fault features of rolling bearings leads to the difficulty in identifying fault types,an approach based on Ensemble Empirical Mode Decomposition(EEMD)and GG(Gath-Geva,GG)clustering is proposed.Firstly,the vibration signal of the bearing was decomposed by the EEMD decomposition method and four intrinsic mode functions(IMF)components containing major fault information were extracted by combining the correlation coefficient principle,and their energy percentage was calculated as the characteristic value,and then the characteristic value was analyzed by GG clustering.The superiority of GG clustering was verified by simulation,and then compared and analyzed the clustering effect of bearing fault data with the GG clustering method proposed in this paper,FCM clustering and GK clustering,and verified the feasibility of the proposed method in fault identification of rolling bearing.

关 键 词：GG聚类 能量比 滚动轴承 

分 类 号：TH16[机械工程—机械制造及自动化] TG506[金属学及工艺—金属切削加工及机床]