单点激振模态参数识别渐近不确定性及实验验证

ASYMPTOTIC IDENTIFICATION UNCERTAINTY OF MODAL PARAMETERS WITH KNOWN INPUT AND ITS EXPERIMENTAL VERIFICATION

  • 摘要: 结构模态参数(频率、阻尼比、振型等)是结构健康监测和动力测试的关键参数,其识别不确定性的准确度量将显著增强结构损伤识别和状态评估的准确性和鲁棒性。着眼于“理解”而非“计算”模态参数识别不确定性,渐近不确定性在不依赖特定数据和算法的条件下给出模态参数估计值变异系数的解析形式。该文介绍单点激振条件下结构的模态参数识别渐近不确定性,基于长数据和小阻尼比假设,给出结构的频率、阻尼比和振型后验变异系数的解析表达式,并结合一系列现场测试进行验证。该文提出的模态参数识别渐近不确定性可用于指导结构激振测试实践,致力于解决振动测试的盲目性和模态识别的被动性。

     

    Abstract: Structural modal parameters (frequencies, damping ratios and mode shapes, etc.) play a critical role in structural health monitoring and dynamic test. Knowing their identification uncertainty can significantly enhance the accuracy and robustness of structural damage detection and condition assessment. Focusing on ‘understanding’ rather than ‘computing’ the identification uncertainty, asymptotic uncertainty gives the analytical forms of the coefficient of variation (c.o.v.) of identified modal parameters regardless of the particular dataset and identification algorithm. This paper introduces the asymptotic identification uncertainty of modal parameters in forced vibration test with known single input. Based on the long data and small damping assumptions, asymptotic expressions for the posterior c.o.v. of structural frequency, damping ratio and mode shape are given. A series of field tests are carried out for verification. The asymptotic identification uncertainty can be used to guide the practice of dynamic testing, to resolve the blindness of the dynamic test and the passiveness of the modal identification.

     

/

返回文章
返回