Cable-stayed Bridge Damage Identification Based On Vehicle-bridge Interaction Vibration Analysis

Traditional health monitoring methods for detecting bridge damage often need to solve complex mathematical inversion problems to identify and locate the damage;compared with traditional structural health monitoring methods,it uses moving vehicles to collect dynamic response data during operation,The health of the bridge can be assessed at a faster speed and with less equipment.Since the vehicle and the bridge are a coupled system,the data collected by the vehicle contains structural information about the bridge,which can be used as a health indicator.This study takes a long-span cable-stayed bridge as the research object and establishes the finite element benchmark model of the cable-stayed bridge,mainly studying the problem of cable-stayed damage identification based on moving vehicles.The main research contents are as follows:(1)The application of the implicit and explicit solution methods in the vehiclebridge coupling finite element solution is introduced.A vehicle-bridge coupling analysis example is used to compare the similarities and differences between the two analysis methods,and the mid-span vertical calculations under the two algorithms are compared.The time history data of vertical displacement,vehicle vertical displacement and vehicle vertical acceleration demonstrate the rationality of using explicit analysis to solve this type of vehicle-bridge coupling system problem,which provides a reference for studying vehicle-bridge coupling analysis problems.(2)Established a bridge full unit model suitable for vehicle-bridge coupled vibration analysis;verified the accuracy of the model by comparing the natural frequency of the bridge in the Midas model.The finite element model of the vehicle was established through Hyper Mesh software,and the three main systems of the vehicle were processed in detail: the suspension system,tire system and connection system;detailed introduction of road unevenness,different vehicle speeds,and vehicle and road contact in simulation experiments Time simulation provides a guarantee for the accuracy of subsequent vehicle-bridge coupling analysis.(3)The relationship between the vertical displacement,the vertical velocity and the vertical acceleration of the bridge and the vehicle and the mutual influence when the stayed cable is non-destructive and damaged,and the vehicle-bridge coupling is analyzed;and then the stay cable is studied The effects of different damage positions and damage levels on structural modalities,bridges and vehicle dynamic response.Finally,vehicle speed and vehicle weight are used as environmental variables to discuss in detail the impact of different vehicle speeds and vehicle weights on damage identification.(4)Introduced the research progress of machine learning,established a CNN neural network algorithm model combined with vehicle acceleration signals for cablestayed bridge damage identification;and took single cable damage identification as an example,focused on 5 Cable element,damage identification under different damage conditions.