LAUSR

The stochastic dynamic damage location vector (SDDLV) method has been widely utilized for localizing damage in structures. However, under changing environmental temperatures, its accuracy is greatly reduced, and the method may even lose efficacy. Furthermore, there has been no investigation of how SDDLV could localize damage in all bridges that are simultaneously monitored within one cluster, i.e., several medium- and small-span bridges with the same structural design on a continuous elevated corridor. To address these issues, a probabilistic SDDLV method is proposed in this study. First, unlike the conventional SDDLV method, the damage location vectors (DLVs) are generated using acceleration monitoring data acquired in the same period that are obtained from each pair of identical bridges among all the monitored bridges in one cluster. Second, considering variations in the finite element model (FEM) of bridges with respect to the environmental temperature, the probabilistic FEM is applied to replace the regular definite FEM, which is utilized to calculate the stresses of all the elements of the structures. Third, a method based on the probability characteristics of the damage localization index is presented to determine the thresholds of damage localization. Then, by incorporating hypothesis testing and a cross-validation strategy, the structural damage of all monitored bridges within one cluster is localized. Finally, a numerical simulation example is utilized to verify the effectiveness of the proposed method, and the effect of structural deviations in construction on the damage localization results is analyzed through the monitoring data of actual bridges.