欢迎访问机床与液压官方网站!

咨询热线:020-32385312 32385313 RSS EMAIL-ALERT
基于经验模态与奇异值分解的振动源数估计方法
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

基金项目:

高档数控机床与制造装备科技重大专项(2017ZX04011001);中国工程物理研究院创新发展基金项目(K1284)


Vibration Source Number Estimation Method Based on EMD and SVD
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    针对振动传感器数小于本底振源数的源数估计问题,提出一种基于经验模态分解的虚拟通道扩展方法。通过经验模态分解得到的固有模态函数构建振动信号观测矩阵,以扩充振动传感器观测通道数量。针对奇异值分解特征矩阵中噪声和数据观测误差以及不确定性导致的源数估计不准确问题,提出一种基于类内散度与类间距离比值优化的聚类分析方法。通过对奇异值分解后特征值矩阵中对角线特征值的聚类分析,获得盲源数估计结果。结果表明:与传统方法相比较,所提方法可准确实现振动信号盲源数估计。

    Abstract:

    A virtual channel expansion method based on empirical mode decomposition(EMD) was proposed to estimate the number of sources when number of vibration sensors was less than number of the vibration sources.The observation matrix of vibration signal was constructed by using the inherent mode function(IMF) obtained through EMD method to expand the number of observation channels of vibration sensor.Aiming at the problems of noise and data observation error in singular value decomposition(SVD) eigenmatrix and the source number estimation inaccuracy caused by uncertainty,a clustering analysis method was proposed based on the optimization of intra class divergence and distance ratio between classes.The blind source estimation result was obtained by cluster analysis of diagonal eigenvalues in the singular value decomposition matrix.The results show that compared with the traditional method,by using the proposed method,the blind source number of vibration signals can be accurately estimated.

    参考文献
    相似文献
    引证文献
引用本文

刘维新,叶超.基于经验模态与奇异值分解的振动源数估计方法[J].机床与液压,2022,50(10):182-187.
LIU Weixin, YE Chao. Vibration Source Number Estimation Method Based on EMD and SVD[J]. Machine Tool & Hydraulics,2022,50(10):182-187

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2022-06-30
  • 出版日期: 2022-05-28