基于PCA的桥梁静态监测数据频域分析方法  

Frequency Domain Analysis Method of Bridge Static Monitoring Data Based on PCA

作　　者：唐浩[1] 梁杨 余建 TANG Hao;LIANG Yang;YU Jian(China Merchants Chongqing Communications Technology Research&Design Institute Co.,Ltd.,Chongqing 400067;Chongqing Municipal Facilities Operation Guarantee Center,Chongqing 400015)

机构地区：[1]招商局重庆交通科研设计院有限公司,重庆400067 [2]重庆市市政设施运行保障中心,重庆400015

出　　处：《公路交通技术》2020年第4期78-82,共5页Technology of Highway and Transport

基　　金：国家重点研发计划项目(2017YFC0806001);重庆市技术创新与应用发展专项重点项目(cstc2019jscx-gksbX0071)。

摘　　要：针对桥梁长期静态安全监测数据量大、影响因素复杂、时域信号难于分解等问题,根据桥梁结构响应信号中各种作用效应的周期重现规律,采用快速傅里叶变换在频域内对多种组成成分进行分离,去除或减小活荷载效应、短时温差效应以及随机干扰误差的影响。利用PCA提取低频段主分量信息用以表征桥梁结构运营状态,为在役桥梁的管养提供科学参考依据。理论推导及实桥案例证明该方法有效,为桥梁安全监测数据分析提供了一种新颖手段。In view of large volume of and complex factors affecting the long-term static safety monitoring data of bridges and the difficulty in decomposing the time-domain signals,a fast Fourier transform is used to separate various components in the frequency domain to remove or reduce the effects of live load effects,short-term temperature difference effects,and random interference errors,according to the periodic reproduction law of various effects in the bridge structure response signals.PCA is used to extract low-frequency principal component information to characterize the operating status of the bridge structure,and provides a scientific reference for maintaining active bridges.The theoretical derivation and actual bridge cases prove the effectiveness of the method,which provides a new method for the analysis of bridge safety monitoring data.

关 键 词：主分量分析 频域分析 桥梁监测 多组分 

分 类 号：U446[建筑科学—桥梁与隧道工程]