Damage Identification of Stay Cables Based on a Small Amount of Deflection Monitoring Data

Abstract

For the problem of cable damage identification in cable-stayed bridges, we have presented a method for identifying cable damage based on deflection monitoring data from a small number of measurement points. We first describe the method to reduce the number of measurement points. We analyzed the distributional characteristics of the deflection difference before and after cable damage in cable-stayed bridges with optimized measurement points. The first derivative of the deflection difference is transformed by a wavelet transform to identify the location of the damaged cable. Then, the Kriging proxy model with exponential and deflection differences is established. The objective function is constructed from the residual deflection difference formed by the deflection difference and the measured deflection difference. With the particle swarm optimization algorithm, the damage parameters in the surrogate model are modified to minimize the objective function, and the damage to the cables is then identified. It is shown that the location of the damaged cable can be identified from the deflection data of a small number of measurement points with small error. The degree of damage can be accurately determined using the surrogate model.