Experimental Study on Continuous Bridge-Deflection Estimation through Inclination and Strain

Abstract

When monitoring structural data, incompleteness is a crucial issue that affects structural health monitoring (SHM). Information on displacement is particularly important for bridge state estimation, but it is difficult to measure. To obtain the required data at any position, a hybrid monitoring (HM) algorithm that combines the finite-element model (FEM) with the monitored data is proposed to extend these data from discrete points to the full structure. The aim of this study is to demonstrate the accuracy and adaptiveness of the algorithm by adopting a complex, large-scale bridge model and considering the modeling error and environmental noise. First, the basic idea and theoretical basis of HM is briefly introduced, and a multitype data-fusion method is proposed to improve the accuracy. Then the experimental equipment, FEM, and updating process are introduced. The influences of the global stiffness error and the boundary condition error are subsequently discussed, showing the algorithm robustness. Finally, the experimental results from two quasi-dynamic loading conditions confirm the HM accuracy using different data sources with high computational efficiency. The superiority of the HM method is also validated by comparing it with some existing methods.