基于ROS的全向移动机器人导航系统设计与仿真
Design and Simulation of Omnidirectional Mobile Robot Navigation System Based on ROS

动态环境复杂多变，机器人如何在动态环境下实现自主移动是一个难题。本文以麦克纳姆轮底盘为对象，基于机器人操作系统(ROS)，设计并实现了开放性好和代码复用率高的全向移动自主导航系统。首先，先对麦轮底盘进行URDF建模，运动学分析，根据得到速度关系重写新的底盘ROS节点；其次结合MOVE_BASE框架，搭建自主导航系统，利用SLAM技术构建二维栅格地图，结合AMCL和路径规划算法的融合导航算法开展自主导航测试。实验结果表明，该方法可以实现机器人的自主移动和避障要求，提升路径规划效果，验证了机器人自主导航的可行性。
Dynamic environment is complex and changeable, so it is a difficult problem for robot to realize autonomous movement in dynamic environment. Based on the Robot Operating System (ROS), this paper designs and implements an omnidirectional mobile autonomous navigation system with good openness and high code reuse rate based on the Mecanum wheel chassis. Firstly, URDF modeling and kinematics analysis were carried out for wheat wheel chassis, and new chassis ROS nodes were rewritten. Secondly, the autonomous navigation system was built by combining MOVE_BASE framework, 2D raster map was constructed by SLAM technology, and autonomous navigation test was carried out by combining AMCL and path planning algorithm. Experimental results show that this method can achieve the requirements of autonomous robot movement and obstacle avoidance, improve the effect of path planning, and verify the feasibility of autonomous robot navigation.