工程科学学报，第37卷，增刊1：72-77,2015年5月 Chinese Journal of Engineering,Vol.37,Suppl.1:72-77,May 2015 DOI:10.13374/j.issn2095-9389.2015.s1.012:http://journals.ustb.edu.cn 基于EEMD形态谱和支持向量机复合的滚动轴承故 障诊断方法 姜万录12》，郑直2，胡浩松2) 1)燕山大学河北省重型机械流体动力传输与控制重点实验室，秦皇岛0660042)燕山大学先进锻压成型技术与科学教有部重点实验室， 秦皇岛066004 ☒通信作者，E-mail:zhengzhi@ysu.cdu.cn 摘要针对滚动轴承的内圈、外圈和滚动体故障提出了一种新的诊断方法，该方法融合了集总经验模态分解(EED)、形 态谱和支持向量机($V)三种方法的优势.首先，利用经验模态分解对滚动轴承故障振动信号进行分解，得到若干个具有物 理意义的内禀模态分量(MF):其次，基于最大能量法筛选出含有故障特征信息最丰富的一个内禀模态分量为故障诊断数据 源：再次，对数据源在选定尺度范围内进行形态谱的提取，从而构造故障特征向量：最后，利用支持向量机对滚动轴承的三种 故障进行诊断.研究结果表明，该方法能够有效地诊断出滚动轴承的三种故障，且具有很高的故障诊新正确率. 关键词集总经验模态分解：形态谱：支持向量机：滚动轴承；故障诊断 分类号TH137:TP277 Fault diagnosis of ball bearing based on EEMD morphological spectrum and support vector machine JIANG Wan-u),ZHENG Zhi),HU Hao-song 1)Hebei Provincial Key Laboratory of Heavy Machinery Fluid Power Transmission and Control,Yanshan University,Qinhuangdao 066004,China 2)Key Laboratory of Advanced Forging Stamping Technology and Science (Yanshan University),Ministry of Education of China,Qinhuangdao 066004.China Corresponding author,E-mail:zhengzhi@ysu.edu.cn ABSTRACT Aiming at fault diagnosis of inner race,outer race and rolling element of ball bearing,a fusion method based on en- semble empirical mode decomposition (EEMD),morphological spectrum,and support vector machine (SVM)was proposed.Firstly, the vibration signal was decomposed by EEMD to get several intrinsic mode functions (IMFs)which have physical meanings.Second- ly,the IMF which was rich in fault features was selected as the data source based on power maximum of IMFs.Thirdly,morphological spectrums in some scales of the IMF were extracted,and then they were adopted as the fault eigenvectors.Lastly,the three faults of ball bearing faults were diagnosed by the use of SVM.The conclusion is that the proposed method can diagnosis the faults of the ball bearing with high accuracy. KEY WORDS ensemble empirical mode decomposition:morphological spectrum:support vector machine:ball bearing:fault diag- nosis 滚动轴承是机械设备中最为常用且举足轻重的零工业和石化工业等重要的生产领域.轴承用于支撑机 部件，它已经被广泛应用到航空航天、工程机械、治金 械设备中的旋转体，减小转动过程中的摩擦，由于工作 收稿日期：20150106 基金项目：国家自然科学基金资助项目(51475405,51075349)：河北省自然科学基金资助项目(E2013203161)工程科学学报，第 37 卷，增刊 1: 72--77，2015 年 5 月 Chinese Journal of Engineering，Vol． 37，Suppl． 1: 72--77，May 2015 DOI: 10． 13374 /j． issn2095--9389． 2015． s1． 012; http: / /journals． ustb． edu． cn 基于 EEMD 形态谱和支持向量机复合的滚动轴承故 障诊断方法 姜万录1，2) ，郑 直1，2) ，胡浩松1，2) 1) 燕山大学河北省重型机械流体动力传输与控制重点实验室，秦皇岛 066004 2) 燕山大学先进锻压成型技术与科学教育部重点实验室， 秦皇岛 066004  通信作者，E-mail: zhengzhi@ ysu． edu． cn 摘 要 针对滚动轴承的内圈、外圈和滚动体故障提出了一种新的诊断方法，该方法融合了集总经验模态分解( EEMD) 、形 态谱和支持向量机( SVM) 三种方法的优势． 首先，利用经验模态分解对滚动轴承故障振动信号进行分解，得到若干个具有物 理意义的内禀模态分量( IMF) ; 其次，基于最大能量法筛选出含有故障特征信息最丰富的一个内禀模态分量为故障诊断数据 源; 再次，对数据源在选定尺度范围内进行形态谱的提取，从而构造故障特征向量; 最后，利用支持向量机对滚动轴承的三种 故障进行诊断． 研究结果表明，该方法能够有效地诊断出滚动轴承的三种故障，且具有很高的故障诊断正确率． 关键词 集总经验模态分解; 形态谱; 支持向量机; 滚动轴承; 故障诊断 分类号 TH137; TP277 Fault diagnosis of ball bearing based on EEMD morphological spectrum and support vector machine JIANG Wan-lu1，2) ，ZHENG Zhi 1，2)  ，HU Hao-song1，2) 1) Hebei Provincial Key Laboratory of Heavy Machinery Fluid Power Transmission and Control，Yanshan University，Qinhuangdao 066004，China 2) Key Laboratory of Advanced Forging ＆ Stamping Technology and Science ( Yanshan University) ，Ministry of Education of China，Qinhuangdao 066004，China  Corresponding author，E-mail: zhengzhi@ ysu． edu． cn ABSTRACT Aiming at fault diagnosis of inner race，outer race and rolling element of ball bearing，a fusion method based on en￾semble empirical mode decomposition ( EEMD) ，morphological spectrum，and support vector machine ( SVM) was proposed． Firstly， the vibration signal was decomposed by EEMD to get several intrinsic mode functions ( IMFs) which have physical meanings． Second￾ly，the IMF which was rich in fault features was selected as the data source based on power maximum of IMFs． Thirdly，morphological spectrums in some scales of the IMF were extracted，and then they were adopted as the fault eigenvectors． Lastly，the three faults of ball bearing faults were diagnosed by the use of SVM． The conclusion is that the proposed method can diagnosis the faults of the ball bearing with high accuracy． KEY WORDS ensemble empirical mode decomposition; morphological spectrum; support vector machine; ball bearing; fault diag￾nosis 收稿日期: 2015--01--06 基金项目: 国家自然科学基金资助项目( 51475405，51075349) ; 河北省自然科学基金资助项目( E2013203161) 滚动轴承是机械设备中最为常用且举足轻重的零 部件，它已经被广泛应用到航空航天、工程机械、冶金 工业和石化工业等重要的生产领域． 轴承用于支撑机 械设备中的旋转体，减小转动过程中的摩擦，由于工作