Structural change monitoring of a cable-stayed bridge by time-series modeling of the global thermal deformation acquired by GPS

Abstract

In this study, the use of ARIMA model coefficients extracted from the response of global thermal deformation, which can be acquired by the GPS monitoring, was proposed for the structural change monitoring of the long-span bridge. The daily periodic air temperature change causes the characteristic global thermal deformation, which is suitable to be acquired by the GPS in the long-span cable-stayed bridge. The pattern of this global thermal deformation was then expected to have sensitivities to changes on global structural properties. The procedures of feature extraction based on the ARIMA model estimation and the Mahalanobis distance comparison were presented, and their applicabilities were verified both by the numerical study and by the application to actual GPS monitoring data. In the numerical study, a FE model of a cable-stayed bridge was constructed, and the time-series displacements under the periodic temperature load were obtained with some cases of structural conditions by varying the boundary conditions and the cable tensions. The procedures of feature extraction and comparison were then applied to the obtained displacement time-series. In the results, the Mahalanobis distance of feature vector, which was configured by estimated AR and MA coefficients, showed significant changes both in the two cases of boundary condition and stayed-cable tension changes. The procedure was then applied to year-round GPS data acquired in the actual cable-stayed bridge. It was shown that the Mahalanobis distance comparison could provide proper assessment to the structural changes that was consistent with the actual structural condition.