基于MED和EEMD的滚动轴承故障诊断方法  

Fault diagnosis of rolling element bearings based on MED and EEMD

作　　者：佘博[1] 田福庆[1] 梁伟阁[1] 

机构地区：[1]海军工程大学兵器工程系,武汉430033

出　　处：《海军工程大学学报》2017年第1期107-112,共6页Journal of Naval University of Engineering

摘　　要：针对滚动轴承早期微弱故障特征难以提取的问题,提出将最小熵反褶积(MED)和集成经验模态分解(EEMD)方法相结合用于提取轴承微弱故障特征的方法。首先,采用MED对滚动轴承振动信号降噪,以增强冲击特征;然后,利用EEMD分解降噪后信号得到一组固有模态分量(IMF),依据相关系数和峭度准则,选择敏感的IMF分量重构信号,并采用希尔伯特包络解调提取故障特征;最后,通过仿真信号和实验台信号验证了该方法的有效性。To solve the problem that it is difficult to extract the early fault signals of rolling bearings, a combination of minimum entropy deconvolution （MED） and ensemble empirical mode decomposition （EEMD） is proposed.The noise can be effectively restrained and the impulse components of vibration signals can be highlighted through the MED method. Then the signals are decomposed into different intrinsic mode function（IMF） by using EEMD. The sensitive IMFS are selected to reconstruct the new signals in accordance with kurtosis and correlated coefficients between IMF and vibration signals. Hilbert transform is used to obtain the weak fault features. The diagnosis results of simulation signals and experimental data of outer ring faults and rolling element faults verify the effectiveness and accuracy of the method.

关 键 词：轴承 最小熵反褶积 集成经验模态分解 相关系数 峭度 

分 类 号：TN911.72[电子电信—通信与信息系统]