BDS/GPS deformation analysis of a long-span cable-stayed bridge based on colored noise filtering

Abstract

Combining GPS and BDS technology to monitor the deformation of long-span railway bridges with stricter deformation control requirements is of significance to the safety and control of the bridge and the safety of railway traffic. Previous studies have ignored the influence of coloured noise in the deformation time series. This is not conducive to accurate deformation analysis of long-span railway bridges. Therefore, GPS, BDS and GPS/BDS monitoring data of Ganjiang Bridge located in Ganzhou city, Jiangxi Province, China are adopted in this paper to filter the coloured noise in the deformation time series by principal component analysis (PCA), and the influence of coloured noise on the deformation analysis results of railway bridge is analysed. The experimental results show that the diurnal temperature difference causes the mid-span and the tower of the railway cable-stayed bridge to deform with a period of about one day in the vertical and longitudinal directions, respectively. Ignoring colored noise will make the uncertainty of the deformation parameter estimation overly optimistic. PCA can significantly reduce the coloured noise, and thus reduce the uncertainty of deformation parameter estimation by about 73%. Moreover, the average difference between the daily periodic motion amplitudes of the monitoring points obtained by using GPS and BDS deformation time series is 1.65 mm. The use of GPS/BDS deformation time series is not only helpful to reduce the influence of coloured noise, but also can reduce the difference between amplitude analysis results obtained from GPS and BDS deformation time series.