Path Planning And Trajectory Tracking Control For Mobile Robot Based On Lidar

In recent years,with the artificial intelligence technology entering the golden age of development,the mobile robotics has also ushered the new opportunities to develop.As a core of the robotics,the technology of SLAM-based indoor autonomous navigation has attracted widespread attention from scholars at home and abroad.With the background of the robot's indoor autonomous navigation,this article aims to the theoretical research on SLAM technology,robot self-localization and path planning,relevant experiments to verify the effectiveness of the designed navigation system,and the practical application combining the distributed network with ROS.Firstly,this paper introduces the hardware configuration of mobile robot in detail,and the overall framework of navigation system is expounded to complete the design of autonomous navigation.Based on the relation among each coordinate system's transformation,the robot system can be modeled.Furthermore,the construction principle of grid map will be analyzed for the subsequent study of the SLAM.Secondly,the arithmetic principle of filter algorithm is researched comprehensively.Meanwhile,the basic algorithm of Rao-Blackwellized Particle Filters is studied by combining particle filter with SLAM.By view of the large computation and the degradation of particles in RBPF-SLAM algorithm,the Gmapping algorithm based on the improved RBPF is selected to set up the indoor map,exploring the optimization measures such as the distribution of improved proposal and the adaptive resampling.Accordingly,with the effects of the simulation platform,the optimal parameters of the Gmapping algorithm will be obtained.Then,this paper introduces the basic principle of adaptive Monte Carlo localization algorithm,and verifies the effect of localization algorithm through simulation experiments.On this basis,by analyzing and comparing the advantages and disadvantages of A* algorithm,LPA* algorithm and D*Lite algorithm,this paper adopts the D*Lite algorithm as the global path planning algorithm of the navigation,using the smooth spline to deal with the path node list.And the path replanning effect of improved D*Lite algorithm is tested with the simulation experiment.In addition,the dynamic window algorithm is adopted to implement the local path planning,which combined with the simulation platform to evaluate the function of the autonomous obstacle avoidance and path tracking,so as to complete the design of the path planning module.Finally,the SLAM experiment is performed in real environment with the mobile robot platform,and the mileage information is optimized with imu data to test the optimization effect of the improved Gmapping algorithm.The effect of robot localization is verified by autonomous navigation experiments,and path planning experiments are carried out respectively in static and dynamic environments.Thus,the ability of path replanning of improved D * Lite algorithm and the local obstacle avoidance of dynamic window algorithm are tested,which prove the reliability of the designed navigation system.