Design Of 2D Localization And Motion Control For Omnidirectional Robots

With the application of robot technology in more and more fields,especially with the development of embedded technology and robot operating system,wheeled robot with ROS system has a wide application background in practical projects.This dissertation focuses on the kinematics model,SLAM mapping and path planning algorithm of omnidirectional robot.In this dissertation,the mathematical model of an omnidirectional robot driven by Mecanum wheels is first established,and the motion control problems in the scene without obstacles and in the environment with obstacles are discussed.In the moving environment without obstacles,the motion control problem of omnidirectional robot can be transformed into a point stabilization control problem.When there are obstacles in the moving environment,the point stabilization problem can be divided into path planning and trajectory tracking control strategies.Secondly,this dissertation studies the mathematical model of lidar,the grid processing of maps and the method of extracting static maps from dynamic maps,studies the robot SLAM algorithm based on Gmapping and Cartographer,implements mapping experiments based on Gmapping and Cartographer in ROS system respectively,and analyzes and compares the two algorithms.Subsequently,this dissertation studies the optimal path planning method of mobile robot in the static grid map constructed,studies the global path planning algorithm based on A * and Dijkstra,and the local path planning algorithm based on DWA and TEB,improves the A * algorithm and proposes a fusion algorithm of global/local path planning.The improved A * algorithm and DWA algorithm are combined to complete the planning of all and local optimal paths respectively,In the ROS simulation environment built in this dissertation,the optimization and security of the fusion algorithm are verified by simulation experiments.Through numerical simulation and experimental discussion,the path planning effect of the location algorithm is illustrated.Finally,in order to verify the effectiveness of the control system designed and implemented in this dissertation,this dissertation uses the omnidirectional wheeled robot experimental platform to build the ROS experimental environment.Based on the ROS system,under the unconstrained conditions without obstacles and the constrained conditions with obstacles,the mapping and autonomous navigation experiments from the unknown environment are completed.The experimental process shows the application effect of the relevant algorithms of mapping and navigation studied in this dissertation.