Research And Implementation Of Obstacle Avoidance Technology For Indoor Mobile Robot

The mobile service robotics covers the core parts of mechanical structure design,autonomous navigation,motion control and artificial intelligence.Autonomous navigation consists of sensor technology,SLAM(simultaneous localization and mapping),adaptive localization and intelligent obstacle avoidance.This project takes the family service robot as the background of the project,and carries out the research work on the intelligent obstacle avoidance related technology of indoor mobile robots.It implements a comprehensive optimal technical solution for the intelligent obstacle avoidance of the robot,aiming at indoor complex and changeable uncertain application scenarios.The research of this paper focuses on static obstacle avoidance technology and dynamic obstacle avoidance technology,which are suitable for indoor mobile robot.Based on the existing obstacle avoidance algorithms,an intelligent obstacle avoidance technical solution is proposed in this paper,which combines A* algorithm,TEB(time elastic band)trajectory optimization algorithm and recovery behaviors,aiming at the security,real-time,stability and accessibility of robot obstacle avoidance.The path search performance of A* algorithm based on grid map is studied and analyzed,and then the cost grid map is processed hierarchically to ensure the security of obstacle avoidance trajectory according to the robot size and the density of obstacle distribution around the robot,which solves the obstacle avoidance security problem caused by the robot map model error,location error,sensor error and so on.In A*algorithm mathematical expressions,the heuristic function is calculated with improved compromise estimation value of Manhattan distance and Euler distance,and deviation cost value is added to the evaluation function,also the redundant path-points are eliminated,thereby the comprehensive performance of this algorithm is much better.What's more,the path planning performance of TEB trajectory optimization algorithm based on multi-objective optimization is studied and analyzed,and the parallel TEB trajectory optimization method based on topological structure is further studied in order to solve the problem of traditional TEB local optimum,and the optimal trajectory is guaranteed by adding the length-cost of trajectory to the evaluation function for the TEB candidate trajectories.While A* pathplanning method is global optimal but lack of real-time.TEB trajectory optimization meets real-time requirements but lack of global information,which means it's easy to fall into local optimal trajectory.This paper combines their advantages in an excellent way that the TEB optimization algorithm find the optimal trajectory in real time under the guidance of global path planned by A* algorithm.The progressive "rotation-clear-replanning" strategy is proposed to restores the robot obstacle avoidance behaviour while obstacle avoidance failure caused by intensive obstacles,no temporary access and so on occurs.Moreover the simulation experiment is done and the whole project is implemented based on ROS(the robot operating system).A mobile robot experimental platform named TURTLEBOT2-RPLIDAR2 consisting of moving base,multiple sensors,operation controller and man-machine interaction client is established.Through the experiments of side-obstacle avoidance,pedestrian avoidance,door moving through and comprehensive obstacle avoidance,it shows that the indoor mobile robot can avoid the obstacles safely timely and reach the target posture quickly and smoothly,which verifies the feasibility of the whole intelligent obstacle avoidance solution.