基于LCD-LLTSA的电动汽车电机轴承故障特征频率提取  

Electric Car Motor Bearing Fault Feature Frequency Extracting Method Based on Local Characteristic-Scale Decomposition and Linear Local Tangent Space Algorithm

作　　者：史素敏[1] 杨春长 王斐 SHI Su-min;YANG Chun-chang;WANG Fei(College of Mechanical and Electrical Information,Shangqiu College,Shangqiu,Henan 476000,China;The 32148 Forces,Zhumadian,Henan 463000,China;First Department,Army Engineering University,Shijiazhuang,Hebei 050003,China)

机构地区：[1]商丘学院机械与电气信息学院,河南商丘476000 [2]32148部队,河南驻马店463000 [3]陆军工程大学一系,河北石家庄050003

出　　处：《计量学报》2020年第10期1267-1272,共6页Acta Metrologica Sinica

基　　金：河北省自然科学基金(E2016506003)。

摘　　要：为有效提取出电动汽车电机轴承故障特征频率,将局部特征尺度分解、线性局部切空间排列和包络分析进行结合,用于电动汽车电机轴承的故障特征频率的提取。首先利用局部特征尺度分解对电动汽车电机轴承故障信号进行分解,得到若干个内禀尺度分量;然后利用线性局部切空间排列对由内禀尺度分量构成的矩阵进行降维处理,得到低维矩阵并以此进行信号重构;最后对重构信号进行包络谱分析,获得故障特征频率。仿真信号和实验信号的实验结果验证了方法的有效性。In order to extracting electric car motor bearing fault feature frequency effectively,a fault feature frequency extracting method of electric car motor bearing based on local characteristic-scale decomposition,linear local tangent space algorithm and envelope spectrum analysis is introduced.Firstly,electric car bearing original fault signals are decomposed into several intrinsic scale component with different frequency band components through local characteristic-scale decomposition(LCD).Then,the linear local tangent space algorithm is used to reduce the dimension of the matrix construct by intrinsic scale component components,and then a new fault signal can obtain by a low dimension matrix which obtained by linear local tangent space algorithm(LLTSA).Finally,the fault frequency can be identified accurately by the envelope spectrum.The experimental results of simulation signal and experiment signal show that the proposed method can identify different state effectively and has a certain superiority.

关 键 词：计量学 滚动轴承 故障诊断 特征频率 局部特征尺度分解 线性局部切空间排列 

分 类 号：TB936[一般工业技术—计量学] TB973[机械工程—测试计量技术及仪器]