随机子空间产生虚假模态及模态遗漏的原因分析

ANALYSIS HOW STOCHASTIC SUBSPACE IDENTIFICATION BRINGS FALSE MODES AND MODE ABSENCE

参数识别是结构健康监测领域研究中的重点。随机子空间法是近年来发展起来的一种线性系统辩识方法,可以有效地从环境激励的结构响应中获取模态参数。由于它具有只需给定系统的阶次一个参数,不存在不收敛的情况等优势,越来越受到国内外业内人士的广泛关注。但是该方法也不是十全十美的,容易产生虚假模态和模态遗漏现象,这些都严重影响了识别效果。因此将识别结果中的虚假模态剔除,是随机子空间方法进一步在理论和应用上拓展的关键。要做到这些,首先要分析虚假模态产生的原因。该文就对虚假模态的产生原因进行了分析。分析表明虚假模态产生的原因主要有两方面:一方面是由于随机子空间方法的基本计算过程而导致的;另一方面是由于实际应用中输入信号不满足白噪声的假定和/或输出信号受到环境的干扰而导致的。分析中采用了一数值模拟例子。

Parameter identification is currently one of the main research topics in the area of structural health monitoring. Stochastic subspace identification is a novel approach developed recent years. It can identify modal parameters of linear structure from ambient vibration of structure. Stochastic subspace identification does not involve any iteration and the only one parameter to be decided is the rank of system. So the method receives more and more attention. But stochastic subspace identification is not perfect, it has some disadvantage such as false modes and mode absence. False modes are the primary ones. These disadvantages distort the identification results. So distinguishing false modes is the key to develop stochastic subspace identification in theory and application. To do this, the first step is to analyze how false modes come into being in stochastic subspace identification. The paper presents the analysis. Research indicates that there are two sources that bring about false modes. One is the algorithm of stochastic subspace identification, the other is that input does not satisfy the assumption of stochastic subspace identification, which assume that input is zero mean white noise and/or output is contaminated by non-white noise. During analyzing, a numerical simulation is adopted.