基于LMD和灰色相似关联度的轴承故障诊断方法  

Roller Bearing Fault Diagnosis Method Based on LMD and Grey Similar Incidence

作　　者：裴峻峰[1] 于志远[1] 董雪[1] 刘威成 

机构地区：[1]常州大学机械工程学院,江苏常州213016

出　　处：《机械设计与制造》2015年第4期66-69,共4页Machinery Design & Manufacture

基　　金：国家自然科学基金项目(51175051)

摘　　要：针对轴承故障振动信号的非线性、非平稳特点及振动信号的强噪声背景,提出一种基于局部均值分解(Local Mean Decomposition,LMD)和灰色相似关联度的轴承故障诊断方法。首先对信号进行局部均值分解,得到若干个PF(Product function,简称PF)分量,再选取包含主要故障信息的PF分量进一步分析,并提取特征向量,然后通过计算标准故障模式与待识别样本的灰色相似关联度对轴承故障类型进行判断。利用该方法对试验轴承故障振动信号进行了分析,结果表明,基于LMD和灰色相似关联度方法能够有效地识别轴承运行状态,实现对轴承的故障诊断。According to nonlinear and non-stationary characteristics and the strong noise background of vibration signals of fault signals for rolling bearings, a method for bearing fault diagnosis is proposed based on local mean decomposition （Local Mean Decomposition, LMD） and gray similarity incidence. Firstly, the signal is decomposed into several PFs component （Product function, referred to as PF） by LMD method, and then selects several PFs component which contains the main fault information f or further analysis and extracts the feature vector. Then the grey incidence of different roller bearing vibration signals is calculated to identify the fault pattern and condition. Using this method, the practical results show that the proposed method can identify the bearing operating state and complete bearing fault diagnosis effectively.

关 键 词：轴承 LMD 灰色相似关联度 故障诊断 

分 类 号：TH16[机械工程—机械制造及自动化] TH133.33