Research On Path Planning Of Wheeled Mobile Robot Based On SLAM Algorithm

In recent years,robotics technology has been widely used in agriculture,medical,military,scientific research and other fields.For an autonomously-mobile robot,how to make it accurately plan a safe and optimal path? How to accurately self-locate and avoid obstacles during its autonomous navigation? The technology of robots’ position-mapping and pathplanning have gradually become the research focus of robot intelligent control.Starting from the theoretical research and engineering application direction,this dissertation takes the self-built wheeled mobile robot as the research object,and studies the map construction and path planning of the wheeled mobile robot based on the lidar sensor.The operating platform of the robot is built by using the SLAM and navigation framework in the ROS system.The platform mainly includes five parts: sensor data,coordinate conversion,SLAM,path planning and motion of the robot.Firstly,in the part of map construction,the mainstream SLAM algorithm is introduced and analyzed in detail.By comparing and analyzing the data obtained by Gmapping and Hector algorithms in Gazebo simulation environment,The Gmapping algorithm with good mapping effect is selected as map-construction algorithm of this dissertation.Secondly,in the part of path planning,the A~* algorithm is improved aiming at the problems of the traditional A~* algorithm,such as serious memory occupation and long calculation time.Based on the A~* algorithm,the idea of jump point is introduced,and a preprocess is added,so as to reduce the unnecessary calculation of A~* algorithm for a large number of intermediate nodes,thus greatly reducing the maintenance of openlist and Closelist,and reducing the memory consumption.A~* algorithm and improved A~* algorithm are simulated and tested on MATLAB and Gazebo.The results show that the improved one with the idea of jump point can plan a safe and collision-free path while ensuring the efficiency of path finding at the same time,greatly reducing the path-finding time.Finally,the corridor on fifth floor and Room 521 of the Third Teaching Building are selected as the test scenarios.The path-planning test of the two algorithm are carried out by using self-built wheeled mobile robot.The analysis of the obtained test data shows that the improved A~* algorithm can generate a better and smooth path,which greatly improves the speed and effectiveness of path finding,thus achieving the expected effect of the experiment.