VMD奇异值和FCM的转子故障特征提取与识别  

Rotor Fault Feature Extraction and Recognition of VMD Singular Value and FCM

作　　者：周萍[1] 姜宏[2] 章翔峰[2] ZHOU Ping;JIANG Hong;ZHANG Xiang-feng(School of Equipment Engineering, Xingjiang Vocation and Technical College of Construction, Xinjiang Urumqi 830047, China;College of Mechanical Engineering of Xinjiang University, Xinjiang Urumqi 830047, China)

机构地区：[1]新疆建设职业技术学院设备工程学院,新疆乌鲁木齐830047 [2]新疆大学机械工程学院,新疆乌鲁木齐830047

出　　处：《机械设计与制造》2018年第6期161-164,共4页Machinery Design & Manufacture

基　　金：新疆维吾尔自治区自然科学基金项目(2015211C256)

摘　　要：为了准确、有效地提取转子故障特征,提出了变分模态分解(VMD)和奇异值特征提取的方法,并采用模糊C均值聚类(FCM)进行转子故障识别。首先,利用分解精度高、模态混叠问题少的VMD算法进行振动信号分解,形成初始特征向量矩阵,然后对该向量矩阵进行奇异值分解,将求得奇异值作为故障特征向量,最后通过模糊C均值聚类形成聚类中心,并计算海明贴近度以实现不同工况下的转子故障分类。将此方法进行转子实验台振动数据验证,实验结果表明:该方法能够有效实现不同工况下转子故障信号的区分,取得了理想的故障诊断结果。In order to extract characteristics of rotor faults accurately and effectively,the paper proposes variarional mode decomposirion（VMD） and extraction approaches for singular value features,and further adopts fuzzy c means clustering（FCM） to identify rotor faults. First,vibration signal decomposition is conducted by applying VMD algorithm of high decomposition precision and less problems in mode aliasing,from which an initial feature vector matrix is formed. And then fault feature vector is acquired by singular value decomposition to the initial feature vector matrix. Finally,form the clustering center by means of fuzzy C-means clustering,calculate simdilar approach degree to realize rotor fault classification under different working conditions. This approach is applied to validate rotor laboratory bench vibration data,the analysis results showed that the method can effectively distinguish the rotor fault signals under different working conditions and achieve the ideal fault diagnosis results.

关 键 词：转子故障 特征提取 VMD FCM 

分 类 号：TH16[机械工程—机械制造及自动化] TH124