基于Ginigram和CHMR的列车轴箱轴承早期故障自主识别方法  

Incipient Fault Autonomous Identification Method of Train Axle Box Bearing Based on Ginigram and CHMR

作　　者：辛格 钟槭畑 李哲 贾利民[3] 杨洋 李林峰 XIN Ge;ZHONG Qitian;LI Zhe;JIA Limin;YANG Yang;LI Linfeng(School of Traffic and Transportation,Beijing Jiaotong University,Beijing 100044,China;Key Laboratory of Transport Industry of Big Data Application Technology for Comprehensive Transport,Ministry of Transport,Beijing Jiaotong University,Beijing 100044,China;State Key Laboratory of Rail Traffic Control and Safety,Beijing Jiaotong University,Beijing 100044,China;CRRC Tangshan Co.,Ltd.,Tangshan Hebei 063000,China;Beijing Institute of Electronic System Engineering,Beijing 100854,China)

机构地区：[1]北京交通大学交通运输学院,北京100044 [2]北京交通大学综合交通运输大数据应用技术交通运输行业重点实验室,北京100044 [3]北京交通大学轨道交通控制与安全国家重点实验室,北京100044 [4]中车唐山机车车辆有限公司,河北唐山063000 [5]北京电子工程总体研究所,北京100854

出　　处：《中国铁道科学》2022年第2期104-114,共11页China Railway Science

基　　金：国家自然科学基金资助项目(51905029);中央高校基本科研业务费专项资金资助项目(2020JBM032,2020JBZD011)。

摘　　要：为解决列车轴箱轴承故障诊断技术面临的多信号源混叠和早期故障程度无法量化的问题,首先,将经典Kurtogram方法采用的峭度改用基尼系数(Gini Coefficient),并将1/3-二叉树滤波方法与基尼系数结合形成Ginigram信号预处理方法,从多源混叠的轴承振动信号中快速提取微弱故障特征;然后,基于预处理信号的平方包络谱,提出一种新的统计指标循环谐波中值比(Cyclic Harmonic-to-Median Ratio,CHMR),有效量化轴承故障部位和程度的信息,并根据西格玛原则完成轴承故障程度的自主分级。为验证方法的有效性,通过列车轴箱轴承试验台进行正常轴承与自然磨损轴承的对比试验。结果表明:相比于采用峭度的经典Kurtogram方法,采用Ginigram对多信号源混叠的故障信号处理效果更优;在故障部位识别方面,CHMR能够精准诊断轴承故障部位;在故障程度量化方面,CHMR相较于现有的循环分量比、二阶循环平稳指标和故障出现率,能更清晰地量化区分轴承的故障程度。To address the issues of multisignal mixing and the inability to quantify the degree of early faults faced for train axle box bearing fault diagnoses,firstly,the kurtosis used by the classical Kurtogram method is replaced by the Gini coefficient.And the Ginigram signal preprocessing method is developed by combining the 1/3-binary tree filtering method with the Gini coefficient to extract weak fault features from the multisource mixing signals of bearing vibration.Then,based on the squared envelope spectrum of the preprocessed signal,a new statistical indicator,Cyclic HarmonictoMedian Ratio(CHMR),is proposed to effectively quantify the location and degree information of bearing faults,and the sigma rule is used to automatically classify the level of bearing fault degree.In order to verify the effectiveness of the proposed method,comparative tests of normal bearings and naturally worn bearings are carried out on the train axle box bearing test rig.The results show that compared with the classical Kurtogram method using kurtosis,the Ginigram is more effective in processing the fault signal with multisignal mixing.In terms of fault location identification,CHMR can accurately diagnose the bearing fault location.And in terms of fault degree quantification,CHMR quantifies and discriminates the bearing fault degree more clearly than the existing ratio of cyclic content,indicator of secondorder cyclostationarity and probability of presence of a fault.

关 键 词：列车轴箱轴承 故障诊断 基尼系数 循环谐波中值比 包络谱分析 

分 类 号：U279.323[机械工程—车辆工程]