Research On SLAM And Navigation Technology Of Indoor Mobile Robot Based On Lidar

With the rapid development of technology,mobile robots have been applied in the fields of industry,service,military,etc.,and gradually entered people's lives.SLAM and navigation technology are the research hotspots of mobile robots.A complete and feasible navigation scheme is very important for mobile robots.For the marketization of home service robots,cost control is necessary.In order to complete the navigation task with low-cost lidar and based on map form SLAM,this thesis studies the SLAM technology and navigation technology separately.With the simulation experiment,a SLAM and navigation system based on low-cost lidar is designed and modularized using the robot operating system ROS.The system has high precision in mapping and navigation,and meets the needs of indoor environment.It is portable and versatile.Firstly,the mobile robot hardware platform and its configuration selected in this paper are introduced.Two mainstream robot operating systems are compared and one of them is initially selected as the robot software platform in this subject.In addition,related models are established for the robot system.The coordinate transformation method and the grid map construction algorithm are introduced to prepare for the subsequent research of SLAM and navigation technology.Secondly,according to different back-end optimization methods,the theoretical research on SLAM technology based on particle filtering and graph-based optimization is carried out.The simulation environment was built in Gazebo,and the simulation experiments were carried out on their representative SLAM schemes gmapping and cartographer.By comparing the mapping effects of the two different schemes,the principle analysis is carried out to select the SLAM scheme which is more suitable for the system.Based on the environmental map,the adaptive Monte Carlo positioning method,the global path planning method based on Dijkstra and A* algorithm,and the local path planning method based on dynamic window search are studied,and the simulation experiment of A* global path planning was carried out in matlab.In the simulation environment built in Gazebo,the simulation experiment based on the map built by SLAM is carried out to verify the path planning,local obstacle avoidance and autonomous navigation.Finally,the overall framework of the mobile robot SLAM and navigation system is designed.In the real laboratory environment,experiments are carried out on the built system,and the mapping results of gmapping and cartographer in real environment are analyzed.Better map was used to try the navigation to the specified target points,and the path planning and dynamic obstacle avoidance capability of the navigation system in real-world environment are verified.The experiments prove that the system has high feasibility and stability,and the navigation accuracy is also high.