一种工作模态参数的在线时域辨识方法

传统模态参数辨识中采用人工激励方法存在着诸多不足。环境激励具有随机性，且线性系统在环境激励下，各输出点响应之间的相关函数与系统的脉冲响应函数具有相同的数学表达式。由此，根据在线模态参数识别的理论，提出了一种依据环境激励响
应识别模态参数的新方法。即用响应谱值法与奇异值分解法来确定系统阶数，解决特征矩阵的定阶问题。该方法简单实用，能在线识别参数，适用于单输入和多输入，但仅适用于激励是白噪声的情况。
关键词：环境激励 模态分析 参数识别
中图分类号：TN830. 7 文献标识码：A
Abstract：Some problems exist in conventional modal parameter identification by manual excitation，involving random environmental excitation
and the same math expressions of correlation function among output point responses and system's pulse response function in the linear system
with environmental excitation. A new method for modal parameter identification depending on environmental excitation response is proposed on
the theory of online modal parameter identification. It can utilize response spectrum value and singular value decomposition to determine system
order number，thus completing character matrix ordering. It features easy-to-use，online parameter identification，and single-input，multi-input
and white noise excitation applications.
Keywords：Environmental excitation Modal analysis Parameter identification