Realization of Multi-point Navigation and Obstacle Avoidance Based on ROS

Abstract

At present, in ROS core framework system, the realization of autonomous cruising and obstacle avoidance is the only way for robots to become more perfect. For this reason, this paper proposes and establishes a physical model of a two-wheel differential-driven mobile robot with nonholonomic motion constraints. SLAM and autonomous navigation are simulated, verified and optimized on STDR and Gazebo platforms, and gmapping and amcl algorithms are selected to realize mapping and navigation on Cater prototype platform. According to the experimental results of simulation verification and optimization, it is shown that using the above platform and algorithm to give the robot an environment map can basically realize accurate positioning and real-time motion planning for positioning targets. In the verification of the prototype, the experimental results are consistent with the theoretical analysis. The feasibility of the method applied in this paper is proved. Keywords-STDR and Gazebo Platforms; Gmapping and Amcl Algorithms; Robot Simulation Navigation; Multi-Point Cruise and Simulation Verification of Intelligent Robot