基于小波域威布尔分布模型的电机滚动轴承故障诊断  

Fault Diagnosis of Motor Rolling Bearing Based on Wavelet-domain Weibull Distribution Model

作　　者：姜海燕[1] 陈苗苗 JIANG Haiyan;CHEN Miaomiao(College of Rail Transit Electric Technology,Hunan Railway Professional Technology College,Zhuzhou 412001,Hunan China)

机构地区：[1]湖南铁道职业技术学院轨道交通电务技术学院,湖南株洲412001

出　　处：《河南科学》2023年第9期1249-1256,共8页Henan Science

基　　金：2018年湖南省教育厅科学研究项目(18C1527)。

摘　　要：电机轴承的运行状态是否正常,通常可以通过分析电机的滚动轴承的振动信号得到诊断结果.因此,分析和研究电机轴承振动信号是滚动轴承研究的热点.首先对电机滚动轴承振动信号进行降噪等预处理,并对预处理的振动信号进行小波分解,再对小波分解系数进行单支重构,得到不同尺度下的单支重构信号;接着对单支重构信号分别建立威布尔分布模型,并验证模型的恰当性,然后求取单支重构信号的威布尔分布模型的尺度参数和形态参数;最后将模型的尺度参数和形态参数输入SVM模式识别器进行故障诊断和模式识别,识别结果表明其参数能较好地表征电机轴承的运行状态.实验结果证明所提方法能较好地诊断电机轴承的故障.We can usually get the diagnosis result whether the running state of the motor bearing is normal or not by analyzing the vibration signal of the motor rolling bearing.Therefore,the analysis and research of motor bearing vibration signal is a hot spot in rolling bearing research.Firstly,the vibration signal of motor rolling bearing is preprocessed by noise reduction,and the preprocessed vibration signal is decomposed by wavelet.Then,the wavelet decomposition coefficient is reconstructed by single branch to obtain the single-branch reconstructed signal at different scales.Secondly,the Weibull distribution model is established for the single-branch reconstructed signal,and the validity of the model is verified.Then the scale parameter and the morphological parameter of the Weibull distribution model of the single-branch reconstructed signal are obtained.Finally,the scale parameters and morphological parameters of the model are input into the SVM pattern recognizer for fault diagnosis and pattern recognition.The experimental results show that the method proposed in this paper can diagnose the fault of motor bearing well.

关 键 词：滚动轴承 威布尔分布 小波变换 故障诊断 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置]