基于盲源分离的结构模态参数识别

PDF(1751 KB)

振动与冲击 ›› 2010, Vol. 29 ›› Issue (3) : 150-153.

论文

基于盲源分离的结构模态参数识别

张晓丹; 姚谦峰

作者信息 +

The Method of Model Parameters Identification Based on Blind Sources Separation

X.D.Zhang; Q.F.Yao

Author information +

文章历史 +

摘要

本文提出一种结构模态参数识别的新方法。该方法以盲源分离理论中基于二阶统计量的AMUSE算法为基础，通过对测试数据Hilbert变换构建分析数据矩阵，通过求解不同时滞下数据协方差矩阵的广义特征值问题实现对结构模态参数的识别。数值算例结果表明，本文提出的方法不仅适用于实模态情况，同时适用于复模态情况，且计算简单，识别精度高，对测量白噪声有很好的鲁棒性。

Abstract

A new method on structure modal parameters identification is developed in the paper. The method is based on the AMUSE algorithm of second order statistic in the blind source separation. It uses Hilbert transformation of the measured data to build analysis data matrices and realizes the structure modal parameters identification by solving the generalized eigenvalue problems of the variance covariance matrix under different time lags. The results of the numerical applications show that this method remains not only valid for the real modes situation but also for the complex modes situation. Further more, the method possesses the advantages of simple calculation, high accuracy of identification and excellent robust for measuring the white noise.

导出引用

张晓丹;姚谦峰. 基于盲源分离的结构模态参数识别[J]. 振动与冲击, 2010, 29(3): 150-153

X.D.Zhang;Q.F.Yao. The Method of Model Parameters Identification Based on Blind Sources Separation[J]. Journal of Vibration and Shock, 2010, 29(3): 150-153

中图分类号： TU311