GPS与加速度计融合桥梁变形信息提取模型研究  

Reconstruction of bridge dynamics using integrated GPS and accelerometer

作　　者：韩厚增[1] 王坚[1] 孟晓林 

机构地区：[1]中国矿业大学环境与测绘学院国土环境与灾害监测国家测绘地理信息局重点实验室,徐州江苏221116 [2]诺丁汉大学土木工程学院

出　　处：《中国矿业大学学报》2015年第3期549-556,共8页Journal of China University of Mining & Technology

基　　金：新世纪优秀人才支持计划项目(NCET-13-1019);江苏高校优势学科建设工程项目(SZBF2011-6-B35)

摘　　要：针对影响高精度全球定位系统(GPS)变形监测的多路径误差和观测噪声,建立基于经验模态分解的系统趋势分离与滤波去噪模型,改正GPS变形序列,构建带有正则化参数的加速度位移重构模型,分析GPS位移本征模态函数分量与加速度重构位移的相关性,通过加速度重构位移高频分量和GPS位移低频分量重构桥梁真实结构响应,达到提取桥梁变形信息的目的.模拟数据分析表明,基于经验模态分解(EMD)的滤波去噪能够有效削弱测量随机误差,不同噪声水平下,加速度重构位移信号均方根误差均小于1 mm,重构位移相似度达到98%以上.利用Wilford悬索桥实测数据进一步验证得出,提出的模型能够提高桥梁位移监测精度及可靠性,最大的桥梁变形达到2cm,提取到桥梁的振动频率为0.16,1.64,2.16及4.51Hz.Global positioning system （GPS） can detect the displacement of the structures, but the GPS positioning result is usually very noisy arid contaminated by the multipath errors, the empirical mode decomposition （EMD） based systematic error separation and noise filter model was developed and applied to GPS time series, the model with regularization factor was built to reconstruct displacement using acceleration measurements, and the high-frequency component and low-frequency component were reconstructed from acceleration and GPS respectively. The simulated trial shows that EMD can effectively eliminate observation noises, the RMS error u- sing acceleration based displacement reconstruction scheme is below 1 mm, and correlation co- efficients are larger than 98% at different noise levels. The trial carried out on the Wilford Bridge was used to validate the proposed approach, it shows that the proposed technique can improve the accuracy of displacement estimates, the bridge response obtained from the esti- mated displacement reaches 2 cm, where the vibration frequencies of the bridge obtained from spectral analysis of deformation series are 0.16, 1.64, 2.16 and 4.51 Hz, respectively.

关 键 词：GPS 加速度 位移重构 TIKHONOV正则化 经验模态分解 多路径误差 

分 类 号：P228.4[天文地球—大地测量学与测量工程]