Research On Autonomous Mobile Platform Of Intelligent Tunnel Inspection Robot Based On Laser SLAM

Tunnel inspection is an important daily work to ensure the normal operation of tunnels.Currently,manual inspections or rail inspection robots are commonly used in tunnel inspections.However,manual inspections are inefficient and rail inspection robots are expensive and difficult to deploy.Based on the key industry chain project of Shaanxi Province,"VR-based Intelligent Tunnel Inspection Robot Key Technology Research",this paper takes the autonomous mobile platform of tunnel inspection robot as the research object,and aims at the problems in the tunnel inspection industry.The autonomous mobile platform of tunnel inspection robot based on laser SLAM(Simultaneous Localization and Mapping)is designed,which could realizes autonomous perception and autonomous navigation in the tunnel.The specific contents are as follows:Firstly,based on the needs of the tunnel inspection robot,the overall design of the tunnel inspection robot was proposed,and the tunnel inspection robot platform was built.Secondly,based on the analysis of the principle and framework of Cartographer,we found there is a problem in loopback constraint when the robot was used in the narrow tunnel.The Lazy Decision method is used to improve the loopback constraint,and the quadratic interpolation method is used to optimize the pose predictor.Thirdly,based on the analysis of the global path planning algorithm Hybrid A*,in order to solve the problem of the Hybrid A* algorithm frequently switching the driving direction in the narrow tunnel,the Dubins curve is introduced,and the improved Hybrid A* algorithm of the end partition search is proposed.Finally,based on the analysis of the local path planning algorithm TEB(Timed On the basis of Elastic Band),in order to solve the problem of rapid and frequent acceleration changes,an improved method with jerk constraint is proposed.Through the test of the SLAM module and the autonomous navigation module,it can be obtained that the improved Cartographer algorithm has better robustness,the improved Hybrid A* algorithm plans a smoother global path;and the improved TEB algorithm makes driving ride comfort has been effectively improved.