Research On MAP Construction And Path Planning Of Indoor Mobile Robot Based On ROS

Robots are widely used in industrial and life service industries.Accurately constructing maps for positioning and path planning are two key points.In view of the poor reusability of the robot software program,the robot of this subject is equipped with a ROS robot operating system to realize map construction and navigation functions.In the hardware selection process,the principle analysis and experiment of the hardware module were carried out.The robot is equipped with a depth camera(RGB-D)sensor and a Lidar sensor to acquire external environmental information,and uses the IMU and the odometer to acquire the robot pose.A differential two-wheel mobile robot based on stm32 microcontroller and industrial computer was built.The stm32 microcontroller is mainly responsible for the acquisition and fusion of wheel mileage data and gyroscope data;the industrial computer is responsible for the processing of depth camera and Lidar data,map information generation and route planning.The industrial computer and stm32 use USB cable to exchange data and instructions between the upper and lower computers through the communication mechanism of ROS.The map is the basis of robot path planning.In the text,the topological map,geometric information map,and grid map are compared,and the grid map is pointed out as the map format in ROS.At the same time,the filtering-based Gmapping algorithm,the Hector-slam algorithm and the cartographer algorithm for graph optimization are introduced respectively.They are all used in this study as the algorithm for slam map construction,and the relevant theoretical derivation is carried out.The subject robot adopts A* and DWA algorithms as the path planning algorithm of the robot.During the path planning and navigation of the robot,it is necessary to move to the target position safely and quickly based on the global path planning and the partial path planning.The Dijkstra and A* global path planning algorithms are compared,and the simulation experiments of the path planning process are carried out on MATLAB.The results verify the efficiency of the A* algorithm.At the same time,the DWAalgorithm of local planning is analyzed.The path planning simulation experiment was also carried out.In order to verify the algorithm,the mapping and path planning experiments were carried out in the virtual simulation environment and the actual scene.T In the actual environment,the map construction experiment was carried out in two different scenes.Scene 1 was a simple closed environment built by the user according to the rules of China RoboCup in 2019.Scene 2 is a complex laboratory environment.The simulation environment is the model established by Gazebo in ROS according to the proportion of scene 1.The robot of this subject carries two sensors,depth camera and radar.In the simulation and actual scene experiments,the depth camera or laser radar is used to verify the algorithm of mapping and the path planning.In the depth camera,only the Gmapping-based mapping algorithm verification experiment was carried out.In the laser radar,the Gmapping algorithm,the Hector-slam algorithm and the cartographer algorithm were used to verify the experiment.The results show that the Lidar's cartographer algorithm has the best mapping accuracy.Then based on the raster map constructed by the s algorithm,the robot performs path planning experiment and verifies the effectiveness of the path planning algorithm.