Research On The Planning And Evaluation Of Multi-robot Cooperative Detection For Cable-stayed Bridges

As China’s highway transportation network continues to improve,the maintenance of bridges is becoming increasingly important.Ensuring the safe operation of bridges is a matter of great importance to people’s livelihoods,and regular inspections and repairs can effectively extend the service life of bridges.However,due to their complex structure,large span,and high detection difficulty,traditional manual detection methods are often unable to comprehensively and effectively detect cable-stayed bridges,resulting in low efficiency,low accuracy,many blind spots,and high risk.To solve this problem,this thesis proposes a multi-robot joint detection technology based on the research of the national key R&D program project “Large-span Bridge Inspection Operation Robot”,which uses drones,wall-climbing robots,cable robots and underwater robots,bridge inspection comprehensive service command vehicles and other bridge inspection robots to carry out coordinated operations in an efficient detection mode,laying the foundation for the realization of intelligent surface inspection and condition assessment of cablestayed bridges.The main research contents and conclusions of this thesis are as follows:(1)Investigate the structural characteristics,typical diseases and their causes of cable-stayed bridges,and propose a cable-stayed bridge inspection plan based on bridge inspection robots in accordance with inspection and maintenance specifications,including robot inspection scope,inspection focus,performance requirements and inspection process.(2)According to the cable-stayed bridge inspection plan of bridge inspection robots,analyze the body structure,spatial positioning technology and bridge disease recognition technology based on machine vision of drones,cable detection robots,wall-climbing robots and underwater robots developed by the project team.The applicability and division of labor of each robot in cable-stayed bridge inspection were studied and tested on actual bridges.(3)Establish a multi-robot detection system according to the robot detection plan and robot technical characteristics.First establish a BIM model of the cable-stayed bridge and import it into the comprehensive bridge inspection platform.Use grid method to establish environmental map and assign detection tasks to bridge inspection robots.Then use path optimization method based on generating tree coverage algorithm to study the detection path planning of drones on cable-stayed bridge stiffening girder and wallclimbing robots on cable-stayed bridge pylon in two cases: single robot and multiple robots.Determine the communication scheme between multiple robots and comprehensive bridge inspection platform and test it.On this basis,the coordinated detection planning of multiple robots was studied and the optimal scheme for coordinated detection of multiple robots on cable-stayed bridges was finally determined.(4)Propose a technical condition assessment method for cable-stayed bridges based on robot detection data.Determine the evaluation criteria for bridge inspection robot detection data.Embed it into the comprehensive bridge inspection platform developed by the project team.According to robot detection data processing analysis,realize condition assessment of cable-stayed bridges based on robot detection surface image data.Finally verify feasibility of this evaluation method through actual cable-stayed bridge detection case.