基于相空间重构与非线性流形的滚动轴承复合故障诊断  

Multi-fault diagnosis for roller bearings based on phase space reconstruction and nonlinear manifold

作　　者：赵洪杰[1] 潘紫微[1] 童靳于[1] 刘燕[1] 

机构地区：[1]安徽工业大学机械工程学院,安徽马鞍山243032

出　　处：《振动与冲击》2013年第11期41-45,共5页Journal of Vibration and Shock

基　　金：国家自然科学基金资助项目(50975003)

摘　　要：针对滚动轴承振动信号的非平稳及非线性特点,提出基于相空间重构与非线性流形的滚动轴承复合故障诊断方法。将滚动轴承一维振动信号重构到高维相空间,计算重构信号协方差矩阵特征值,以此组成轴承故障诊断原始特征集;采用局部切空间排列算法对原始特征集作特征压缩后,将所得新特征输入到K-means分类器中进行轴承故障识别与聚类。实验结果表明,与经典线性分析方法 PCA相比,该方法聚类效果更好。A multi-faults diagnosis approach for roller bearings based on space reconstruction and nonlinear manifold was proposed according to the fact that vibration signals for roller bearings are non-stationary and time-varying. After embedding a vibration signal into a higher dimensional phase space, its original feature set was acquired by calculating eigenvalues of its covariance matrix. Using the local tangent space alignment algorithm to compress the original feature set, the new features obtained were input into a K-means classifier, and the output of the K-means classifier was clustering results. The experimental results showed that the proposed method has a better clustering performance than the traditional linear principal component analysis (PCA) method does.

关 键 词：滚动轴承 相空间重构 流形 复合故障 局部切空间排列算法 

分 类 号：TH133[机械工程—机械制造及自动化] TP274[自动化与计算机技术—检测技术与自动化装置]