基于随机子空间识别的损伤试验研究  

EXPERIMENTAL STUDY ON DAMAGE MONITORING BASED ON STOCHASTIC SUBSPACE IDENTIFICATION

作　　者：杜永峰[1] 张博[1] 李慧[1] 周晶[1] 

机构地区：[1]兰州理工大学防震减灾研究所,兰州730050

出　　处：《低温建筑技术》2008年第4期52-55,共4页Low Temperature Architecture Technology

基　　金：甘肃省建设科技攻关基金资助项目(JK-2004-18)

摘　　要：采用附加质量模拟损伤的方法通过对三层空间钢框架模型模态试验分析,采用随机子空间(SSI)法进行模态参数的识别,并在模拟健康状态和有损伤状态下识别结果与有限元模态分析及经典的特征值实现法(ERA)结果进行多次对比,对比结果表明了随机子空间法识别模态参数的可靠性。依据随机子空间法识别的模态振型,结合模态置信准则损伤判别的方法,进行了损伤的诊断,诊断的结果判断了结构的损伤程度,搜索出了损伤的最不利工况。Through the modal experiment analysis of three - tier space steel frame model using of adding mass simulated structure damages, it is made use of identifying modal parameters based on stochastic subspace identification and simultaneity the twice contrast with the finite element modal analysis and experiment based on ERA identification method in the different eases which is one of healthy statement, other is simulated damage statement. It is shown that this method identifying modal parameters is reliability. Based on the identified modal shapes and Modal Assurance Criterion （MAC） damage identification method, the paper also judged damage degrees and search the worst condition effectively.

关 键 词：参数识别 随机子空间 模态置信准则 

分 类 号：TU375.4[建筑科学—结构工程]