基于Cohen类时频分布的机械故障源分离  

Blind Separation of Machine Fault Sources Based on Cohen Class Time-frequency Analysis

作　　者：吕亚平[1,2] 李志农[1,2] 曾宇冬[1] 

机构地区：[1]南昌航空大学无损检测技术教育部重点实验室,江西南昌330063 [2]郑州大学机械工程学院,河南郑州450001

出　　处：《机床与液压》2010年第7期134-137,共4页Machine Tool & Hydraulics

基　　金：国家自然科学基金资助项目(50775208);河南省教育厅自然科学基金资助项目(2008C460003;2006460005)

摘　　要：针对传统的机械故障源分离方法忽略信号非平稳性的不足,结合Cohen类时频分布和盲源分离各自的优点,提出一种基于Cohen类时频分析的机械设备非平稳信号盲分离方法,并通过均方根误差比较基于各种时频分布的盲源分离算法的分离性能。仿真和实验结果表明,提出的方法是有效的,机械设备非平稳信号的盲分离必须充分利用信号的非平稳性。The nonstationarity of fault signals is usually neglected in the traditional blind source separation of machine faults. Aimed at this deficiency, combining the advantage of Cohen class time-frequency distributions and blind source separation （ BSS）, a blind separation method of non-stationary signals in the mechanical equipment based on Cohen class time-frequency distributions, which was named Cohen-BSS method, was proposed. The performances of bind separation methods based on several different quadratic time-frequency distribution were compared, separation effect were reflected by the root-mean-square error. The simulation and experimental results show that the proposed method is effective, the non-stationarity must be fully considered in the separation of the machine faults.

关 键 词：盲源分离 非平稳信号 时频分布 故障诊断 

分 类 号：TH17[机械工程—机械制造及自动化] TP911[自动化与计算机技术]