多输入多输出桥梁结构时变模态参数的滑窗时域识别  

Sliding Window Time-domain Method for Identifying Time-varying Modal Parameter of Multi-input and Multi-output Bridge Structure

作　　者：周筱航 单德山[2] Inamullah Khan 李乔[2] ZHOU Xiaohang;SHAN Deshan;Inamullah KHAN;LI Qiao(College of Civil Engineering and Architecture,Guangxi University,Nanning 530004,China;School of Civil Engineering,Southwest Jiaotong University,Chengdu 610031,China;School of Civil Engineering,National University of Sciences and Technology,Risalpur 24080,Pakistan)

机构地区：[1]广西大学土木建筑工程学院,广西南宁530004 [2]西南交通大学土木工程学院,四川成都610031 [3]国立科技大学土木工程学院,巴基斯坦里沙浦24080

出　　处：《应用基础与工程科学学报》2020年第5期1145-1157,共13页Journal of Basic Science and Engineering

基　　金：国家重点基础研究发展计划(2013CB0363);国家自然科学基金项目(51678489);国家重点研发计划(2016YFC0802202)。

摘　　要：为了跟踪识别多输入多输出桥梁结构的时变模态参数,提出了一种基于滑窗技术与确定-随机子空间识别的时域识别方法(Sliding Window-Combined Determine-stochastic Subspace Identification,SW-CDSI).首先,使用滑窗技术对连续测试数据进行窗口划分,讨论了窗函数、窗宽度及滑窗步长的确定标准,提出了基于稳定图法的窗宽度确定方法.然后,使用考虑多输入多输出的确定-随机子空间方法对各窗口内数据进行处理,实现在时域上对桥梁模态参数进行跟踪识别.最后,基于大比尺斜拉桥模型振动台试验测试数据,对SW-CDSI方法进行验证,并将SW-CDSI辨识结果与希尔伯特-黄变换(Hilbert-Huang Transform,HHT)辨识结果进行对比分析.结果表明:SW-CDSI方法可精确地跟踪桥梁频率在时域上的变化情况,其比HHT方法具有更高的精度和稳定性.In order to track the time-varying modal parameters of the bridge structure with multi-input and multi-output,a time-domain identification method based on sliding window technique combined with determine-stochastic subspace identification(SW-CDSI)was proposed.At First,the continuous measurement data was divided into separate windows by adopting sliding window technique,then the determination standard of window function,window length and step length were discussed;and then a window length determination method based on stabilization diagram was proposed.Furthermore,the combined determine-stochastic subspace identification method that could incorporate multi-input and multi-output was applied to cope with the data in each window,thus the tracking of bridge modal parameters in time-domain was realized.At last,SWCDSI method was verified based on the measured data of shaking-table test of a large-scale cable-stayed model bridge;and the identification result of SW-CDSI was compared with that of Hilbert-Huang Transform(HHT).The result indicates that the SW-CDSI can precisely track the frequency variations of bridge in time domain,and it exhibits higher accuracy and stability than HHT.

关 键 词：桥梁 时变 模态参数 滑窗 确定-随机子空间 希尔伯特-黄变换 振动台试验 

分 类 号：U446[建筑科学—桥梁与隧道工程]