特长钢箱梁桥GPS监测数据的奇异谱分析

Singular spectrum analysis of GPS monitoring data on specialty steel box girder bridge

实时准确地监测和分析大型建筑物的变形对于大型建筑物的施工与运行是十分重要的。奇异谱分析(SSA)是一种与经验正交函数相关联的一种统计技术。文中采用奇异谱分析(SSA)的方法,结合功率谱峰值与原序列和RC1-2重建序列均方误差最小的方法确定最优潜入维数M,研究了特长钢箱梁桥索塔的变形趋势和震荡周期。小波分析技术也可以很好地描述时间序列的时频分布,但SSA重构后的结果比小波要平滑些,更能反映出时间序列的特征。结合苏通大桥北索塔监测序列表明,北索塔站的N,E,U方向上均存在明显的趋势和显著的变动周期,而且也含有较多的噪声信息。

Real-time accurate monitoring and analysis on deformation of large buildings for construction and operation are very important.Singular spectrum analysis(SSA)is a kind of statistical technique associated with the empirical orthogonal function.This paper adopts the method of singular spectrum analysis(SSA),combined with a method of power spectrum peak rate with the original sequence and RC1-2reconstruction sequences of minimum mean square error to determine the optimal dimension M.The deformation trend and the oscillation cycle of cable tower are studied for specialty steel box girder bridge.Wavelet analysis technology can well describe the time sequence of the time-frequency distribution.But SSA reconstruction result is more smoothing than wavelet analysis,which can reflect the characteristics of time series better.North cable tower of the Sutong bridge monitoring sequence shows that N,E and U direction of north cable tower stay with obvious trends and significant changes in the cycle,and also contain more information noise.