基于EEMD和RBFNN的列车滚动轴承故障诊断研究  

Research on fault diagnosis of train rolling bearing based on EEMD and RBFNN

作　　者：李笑梅[1] 贺德强[1] 谭文举 陈二恒 LI Xiaomei;HE Deqiang;TAN Wenju;CHEN Erheng(College of Mechanical Engineering, Guangxi University, Nanning 530004, China;Nanning Rail Transit Group Co., Ltd., Nanning 530004, China)

机构地区：[1]广西大学机械工程学院,广西南宁530004 [2]南宁轨道交通集团有限公司,广西南宁530004

出　　处：《铁道科学与工程学报》2017年第5期1056-1062,共7页Journal of Railway Science and Engineering

基　　金：国家自然科学基金资助项目(51165001);广西科技攻关资助项目(1598009-6);南宁市科技攻关资助项目(20151018)

摘　　要：针对列车滚动轴承振动信号的非高斯、非平稳性特征,提出一种基于集合经验模式分解(Ensemble Empirical Mode Decomposition,EEMD)和径向基函数神经网络(Radial Basis Function Neural Network,RBFNN)相结合的滚动轴承故障诊断方法,利用EEMD方法对振动信号进行分解,得到前8个本征模态函数(Intrinsic Mode Functions,IMF)分量,将归一后的IMF能量特征向量作为RBF神经网络的输入向量构建故障诊断模型,从而实现滚动轴承的故障识别。将RBF神经网络方法和BP(Back Propagation)神经网络进行对比,本文提出的方法能精确识别正常轴承、滚动体故障、外圈故障和内圈故障等4种轴承状态,为提高列车滚动轴承故障诊断的准确性和实时性提供了新思路。For the non-Gaussian and non-stability of the vibration signal of train rolling bearing,a method of rolling bearing fault diagnosis is put forward based on Ensemble Empirical Mode Decomposition(EEMD)and Radial Basis Function Neural Network(RBFNN).The vibration signal is decomposed using EEMD method to obtain first eight Intrinsic Mode Functions(IMF)components,and then IMF energy feature vectors are used as the input vectors of the RBF neural network to build a fault diagnosis model,in order to realize rolling bearing fault identification.Finally,compare the RBF neural network and the Back Propagation(BP)neural network.The simulation results show that the proposed method can accurately identify four kinds of bearing states,such as normal bearing,rolling body fault bearing,outer race fault bearing and inner race fault bearing,a new idea is provided to improve the accuracy and real-time performance of the fault diagnosis the rolling bearing for train.

关 键 词：列车 滚动轴承 故障识别 EEMD RBF神经网络 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术] U260[自动化与计算机技术—计算机科学与技术]