Research On Dynamic Path Planning Of Mobile Robot Based On Laser SLAM

Robot autonomous navigation includes simultaneous localization and mapping(SLAM)and autonomous path planning,it is an important part of the research on intelligent mobile robots.In order to improve the robot's ability to plan paths in a dynamic environment,research on dynamic path planning of mobile robot based on laser SLAM in a dynamic environment is carried out in this thesis,it has theoretical significance and practical value.First of all,in this thesis,the research status of autonomous navigation of mobile robot at home and abroad is analyzed.The navigation system of mobile robot based on laser SLAM is designed,lidar and odometer are selected to collect information,Fast SLAM algorithm is selected as the foundation to realize SLAM function,Robot Operating System(ROS)is selected as the software platform.In order to realize the dynamic path planning function,the current main path planning methods are compared,and the D*lite algorithm is selected as the basic algorithm for path planning.Because there are some problems in the D*lite algorithm that reduce the search efficiency,the research focus of this thesis is to improve the D*lite algorithm.During the initial path planning of the D*lite algorithm,if the search space is divided into multiple smaller free areas by obstacles,the correct search direction is hidden,the calculation times are increased,and the search efficiency is reduced,to solve this problem,an improved D*lite initial path planning algorithm based on cell decomposition is proposed.New decomposition rules are added to Boustrophedon cellular decomposition,this new method decomposes the map into multiple units,a graph with these units as nodes is constructed.A bidirectional graph search algorithm is designed to quickly calculate which units the path needs to pass in turn.The main grids are set up in these units and form a linked list in order,this linked list guides the right search direction and makes planning faster.Experimental results show that the improved algorithm improves the efficiency of initial path planning in the initial map.In a dynamic environment,the initial path may be blocked by new obstacles during the robot's movement.At this time,the path needs to be re-planned.D*lite algorithm only needs to re-plan the path in local area,however,the algorithm reinitializes the affected area by operating the priority queue multiple times,the efficiency ofre-planning is affected.To solve this problem,an improved D*lite path re-planning algorithm based on search tree cutting is proposed.The search tree can be constructed by the parent pointer after the initial path planning is completed,affected grids are quickly reinitialized by cutting affected branches,the times of operation on the priority queue are reduced,and the time consumption is reduced.Experimental results show that the improved algorithm can improve the efficiency of path re-planning in dynamic environments.Finally,based on the Fast SLAM algorithm and the path planning algorithm proposed in this thesis,the navigation system of mobile robot based on laser SLAM is built by designing hardware and software.Navigation experiments in a dynamic environment are used to verify the effectiveness of the path planning algorithm proposed in this thesis.Experimental results show that the path planning algorithm proposed in this thesis can effectively complete the path planning of intelligent mobile robots in a dynamic environment,and the navigation system can stably complete the navigation of mobile robots.