Design And Implementation Of SLAM System For Mobile Robot Based On ROS

In today's society,mobile intelligent robots have been applied to all aspects of life,intelligent appliances,mobile navigation and many other industries have mobile robots.A robot that can achieve rapid and accurate positioning is also the mainstream direction of mobile robot research.The main research content of this paper is the design and implementation of SLAM(Simultaneous Localization and Mapping)system of mobile robot based on ROS.The research direction is based on the indoor scene of robot mapping and positioning,as well as robot mobile obstacle avoidance.The slam system of mobile robot based on ROS is designed in this paper,which mainly uses the algorithm of monocular vision fusion IMU to estimate the position and attitude of the robot and improve the accuracy of the system positioning.Through the analysis of project requirements,the project mainly includes software system development and hardware platform construction.The software system mainly includes four functional modules: system initialization,tracking and positioning,map building and obstacle avoidance.Among them,the function of system initialization is to read configuration files and process images before the main program runs,and connect with the interface provided by ROS as the entry of the program the data obtained by sensors to estimate the mobile position and attitude of the robot,so as to track and locate in real time;the map module is built to draw the machine The human environment map and moving track,and optimize its position and posture;the obstacle avoidance module sends a moving command to the robot according to its environment,and controls the robot to move and bypass the obstacles in front.The hardware platform consists of three parts: hardware equipment integration,camera calibration and ROS platform debugging.Hardware equipment integration is to connect the robot,camera,computer and other equipment used in the experiment,and adjust the program to make it run normally;camera calibration is to calibrate the internal and external parameters of the camera,so as to prepare for the normal use of the camera in the future;ROS platform debugging is to adjust the bottom operating system to make the robot run,so as to connect the hardware equipment and software system Middleware.The project of this paper is an experimental project,which is mainly to establish a visual slam system integrating IMU data and build the corresponding experimental hardware platform.The software system obtains the hardware equipment data to locate and map,and uses the software system to control the robot movement,so as to achieve the purpose of real-time positioning and map building by combining hardware equipment and software system.