Beam Bridge Damage Identification by Data Fusion of Accelerometers and Computer Vision Measurements of Moving Vehicle-induced Responses
Damage identification methods based on structural mode-related parameters require a high spatial resolution of the mode shapes.To address the shortcomings of the existing computer vision technology for the full-field mode shape extraction and damage identification of bridge structures and to effectively utilize the abundant data collected by various sensors in the structural health monitoring(SHM)system of bridges,this study proposes a damage identification approach based on computer vision and measurement data of moving load-induced responses from a limited number of accelerometers.First,through theoretical derivation,the displacement responses of a single-span beam under arbitrary boundary conditions subjected to a moving concentrated force before and after damage occurrence were investigated.Subsequently,the changes in structural displacement responses and mode shapes induced by local stiffness damage were analyzed from a theoretical perspective.Furthermore,a method was introduced to extract high-resolution mode shapes,which represent the true state of the structure,from computer vision-based displacements and accelerations measured at specified locations.The spatial resolutions were determined based on the speed of the moving load and the displacement sampling frequency.According to the characteristic of the extracted mode shapes,a damage identification index based on the areal difference of multi-order full-field modal curvatures before and after damage was proposed.This index can be directly calculated based on the mode rotations before and after damage occurrence.To validate the proposed method,a series of laboratory experiments were conducted on a simple supported beam model under various conditions of moving loads with different speeds and amplitudes.The results indicate that the proposed method extracts full-field modes of the model beam through computer vision and a limited number of acceleration sensors.The accuracy of the modes matches well with the results obtained from the mode analysis of the accelerations,and the spatial resolutions are about 1/10000～1/2500 of the span length of the model beam.Moreover,the proposed damage identification index accurately locates the damage under different conditions,even when there are variations in the size and speed of the moving loads before and after damage occurrence.
NETL
NSTL
万方数据
