基于近邻元分析的滚动轴承故障诊断方法

PDF(1634 KB)

振动与冲击 ›› 2015, Vol. 34 ›› Issue (2) : 138-142.

论文

基于近邻元分析的滚动轴承故障诊断方法

周海韬1,2，陈进1,2，董广明1,2

作者信息 +

Fault diagnosis method for rolling bearing based on neighborhood component analysis

ZHOU Hai-tao1,2，CHEN Jin1,2，DONG Guang-ming1,2

Author information +

文章历史 +

摘要

提出基于近邻元分析（Neighborhood Component Analysis, NCA）的滚动轴承故障诊断模型。利用NCA技术对滚动轴承信号的时、频域特征向量降维，并对降维后向量分类，成功区分滚动轴承四种状态。通过Fisher判别函数定量分析目标维数对NCA降维效果影响，确定最佳特征约简目标维数。为突出NCA方法优势，将NCA与PCA两种不同降维方法进行对比。实验结果表明，NCA作为监督式降维方法，其聚类效果好于PCA。

Abstract

This paper proposed a novel fault diagnosis method for rolling bearing based on neighborhood component analysis(NCA). NCA is used to reduce the dimensions of input time and frequency domain features. Then the classification is performed. Fisher evaluation function is applied to select the proper object dimension. In order to show the advantages of the proposed method, the classification results based on both PCA and NCA are compared. The experiment shows that as a supervised dimension reduction method, NCA performs better than PCA.

导出引用

周海韬;陈进;董广明;. 基于近邻元分析的滚动轴承故障诊断方法[J]. 振动与冲击, 2015, 34(2): 138-142

ZHOU Hai-tao;CHEN Jin;DONG Guang-ming;. Fault diagnosis method for rolling bearing based on neighborhood component analysis[J]. Journal of Vibration and Shock, 2015, 34(2): 138-142