一种新的轴承故障诊断方法  

A new method of bearing fault diagnosis

作　　者：张荣涛 焦斌[1] 李彬彬[1] ZHANG Rongtao;JIAO Bin;LI Binbin(School of Electrical Engineering,Shanghai Dianji University,Shanghai 201306,China)

机构地区：[1]上海电机学院电气学院

出　　处：《上海电机学院学报》2019年第5期270-275,共6页Journal of Shanghai Dianji University

摘　　要：针对轴承时域参数中正常数据与故障数据区分困难的问题,将用于提取图像局部纹理特征的局部二值模式(LBP)算法引入到轴承的故障诊断中来,表征中心幅值在应属窗口中的重要性,通过样本熵求取其中新模态出现的概率,提出LBP样本熵的概念,将提取出的样本熵与其他人工提取的特征元素组成特征向量,利用XGBoost算法判断轴承的运行状态。将LBP样本熵与人工参数进行对比,结果表明,LBP样本熵能够轻易地区分正常状态和故障状态,与XGBoost算法相结合对轴承的故障进行识别,相比传统的人工提取特征在轴承时域故障识别中起到了一定的积极作用,与其他故障诊断模型相对比说明LBP样本熵与XGBoost算法的结合具有一定的优越性。In order to solve the problem that it is difficult to distinguish the normal data from the fault data in the time- domain parameters of the bearing,the local binary pattern(LBP) algorithm,which is used to extract the local texture features of the image,is introduced into the fault diagnosis of the bearing to characterize the importance of the central amplitude in the corresponding window.Then,the probability of the emergence of the new mode is calculated by the sample entropy, the concept of the LBP sample entropy is put forward,and the feature vector is composed of the extracted sample entropy and other artificially extracted feature elements.Next,the eXtreme gradient boosting(XGBoost ) algorithm is used to judge the running state of the bearing.The entropy of the LBP sample is compared with the artificial parameter.The results show that the LBP sample entropy can easily distinguish the normal state from the fault state.Finally,the LBP sample entropy is combined with the XGBoost algorithm to identify the bearing faults.The results show that the LBP sample entropy plays a positive role in the bearing time-domain fault recognition compared with the traditional manual extraction features.Compared with other fault diagnosis models,the combination of the LBP sample entropy and the XGBoost algorithm has certain advantages.

关 键 词：故障诊断 时域参数 局部二值模式(LBP) 样本熵 

分 类 号：TH133.3[机械工程—机械制造及自动化] TP277[自动化与计算机技术—检测技术与自动化装置]