增强MCKD和增强EWT的轴承滚珠微弱故障检测  

Bearing Rolling Ball Weak Fault Detection Based on Enhanced Mckd and Enhanced Ewt

作　　者：陈小怡[1] 魏圣坤[1] CHEN Xiao-yi;WEI Sheng-kun(College of Intelligent Manufacturing and Automotive Engineering,Luzhou Vocational and Technical College,Luzhou 646000,China)

机构地区：[1]泸州职业技术学院智能制造与汽车工程学院,泸州646000

出　　处：《组合机床与自动化加工技术》2022年第12期62-66,71,共6页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：智能控制与电子器件应用技术泸州市重点实验室资助项目(ZK202204)。

摘　　要：针对噪声环境下轴承滚珠振动信号传递路径长、特征微弱等特点,提出一种增强经验小波变换(EEWT)和增强最大相关峭度反卷积(EMCKD)相结合的轴承滚珠微弱故障检测方法。首先,针对最大相关峭度反卷积(MCKD)参数难以选取的问题,提出一种功率谱熵评价准则的MCKD参数自适应选取方法,使用网格搜索法对参数进行寻优,利用EMCKD对信号进行初次降噪;其次,针对经验小波变换(EWT)的过分解问题,提出一种信号频谱有效边界划分方法,利用EWT将信号分解为调幅-调频分量(AM-FM);最后,利用分量筛选指标对主要AM-FM分量进行重构并进行功率谱分析,从而实现滚动轴承滚珠微弱故障故障检测。试验结果表明,EMCKD-EEWT方法可有效提取滚动轴承滚珠的微弱故障特征,准确检测滚珠故障。Aiming at the characteristics of long transmission path and weak characteristics of bearing rolling ball vibration signals in noisy environment,a method for weak fault detection of bearing rolling ball was proposed,which combined enhanced empirical wavelet transform(EEWT)and enhanced maximum correlation kurtosis deconvolution(EMCKD).Firstly,in order to solve the problem that the maximum correlation kurtosis deconvolution(MCKD)parameters were difficult to select,an adaptive parameters selection method of MCKD based on the power spectrum entropy evaluation criterion was proposed.The grid search method was employed to optimize the parameters then the first noise reduction was carried out.Secondly,for the over-decomposition problem of empirical wavelet transform(EWT),a method for dividing the effective boundary of signals spectrum was proposed,and the signals were decomposed into amplitude modulation-frequency modulation components(AM-FM)by using EWT.The index reconstructed the main AM-FM components and conducted power spectrum analysis,so as to realize the weak fault detection of rolling bearing balls.The test results show that the EMCKD-EEWT method can effectively extract the weak fault features of the rolling bearing balls and accurately detect the ball faults.

关 键 词：轴承滚珠 经验小波变换 最大相关峭度反卷积 故障检测 

分 类 号：TH133.3[机械工程—机械制造及自动化] TG502[金属学及工艺—金属切削加工及机床]