基于改进EMD-BSS算法的地铁齿轮箱单通道轴承微弱故障特征提取  

Single-channel Metro Gearbox Bearing Weak Fault Feature Extraction Based on Improved EMD-BSS Algorithm

作　　者：白永亮[1,2] 杨建伟[1,2] 姚德臣[1,2] 曾恒 

机构地区：[1]北京建筑大学机电与车辆工程学院,北京100044 [2]北京建筑大学城市轨道交通车辆服役性能保障北京市重点实验室,北京100044 [3]贵阳市城市轨道交通有限公司,贵阳550091

出　　处：《机械设计与研究》2016年第6期57-61,64,共6页Machine Design And Research

基　　金：轨道交通控制与安全国家重点实验室自主课题(RCS2010ZZ002);城市轨道交通车辆服役性能保障北京市重点实验室引导性课题(06080915001);北京建筑大学科学研究基金资助(00331615015)

摘　　要：地铁齿轮箱振动情况复杂,微弱故障特征常被噪声湮没而难以提取。本文提出一种改进的EMDBSS算法,通过计算IMF分量的相关系数矩阵,优化单通道EMD-BSS方法中的IMF信号重组过程,并对重组后的信号进行微弱故障特征提取。将该算法应用于仿真信号及实测地铁齿轮箱轴承故障振动信号中,成功提取出微弱故障特征,验证了算法的有效性。该算法的提出对于单通道微弱故障特征提取具有积极意义。The vibration of metro gearbox is complex, and its weak fault features are difficult to be extracted because of the noise annihilation. This paper presents an improved EMD-BSS algorithm to optimize the IMF restructuring process in single-channel EMD-BSS method by, calculating the correlation coefficient matrix of IMF components. The weak fault feature extraction of the signal after the ~re^rganization is conducted. The algorithm is applied to the simulated signal and the real subway gearbox bearings fa^ult ~ibi-ation^signals, and the weak fault features are extracted successfully, which proves the effectiveness of the algori!hm. The proposed algorithm has positive significance for extracting weak fault features.

关 键 词：地铁齿轮箱 微弱故障 特征提取 单通道EMD-BSS 

分 类 号：TH17[机械工程—机械制造及自动化]