Research On Indoor Localization And Autonomous Navigation Of Mobile Robot Based On Lidar

In recent years,mobile robots are more and more widely utilized in industrial production and people’s life,which has made great contributions to the development of society.With the increasingly complex application environment,it’s particularly important to improve the autonomous ability of mobile robots.SLAM and path planning are the core technologies of autonomous navigation for mobile robots and they have become the research focus in this field.This paper researches the system of localization and navigation,SLAM technology and path planning technology.The tracked mobile robot is used as the carrier to complete the related research verification work.The main contribution of this paper are as follows:In the construction of localization and navigation system of tracked mobile robot,the hardware system of the tracked mobile robot is modeling,and the selection and principle of the hardware equipment are explained in detail.The software framework of the mobile robot is built in ROS,the overall navigation architecture of the tracked mobile robot is determined,and the 3D model of the tracked mobile robot is drawn and imported into the indoor environment built by the simulation software Gazebo.The coordinate system model,the two-wheel differential kinematics model and the odometer model of the mobile robot are theoretically derived in detail.In terms of indoor map construction of mobile robots,the Gmapping,Karto and Cartographer algorithm are researched in detail.For the hardware adaptation of Cartographer algorithm to lidar sensors in practical work,parameter optimization of local SLAM of the algorithm is proposed.The functional principles of voxel filtering parameters,adaptive voxel filtering parameters and subgraph parameters are analyzed.After adjusting the parameter values,the algorithm is run in the recorded lidar data set.The algorithm results are analyzed according to the error evaluation criteria,and the optimal parameter values suitable for the experimental equipment are selected.Aiming at the problems that A*algorithm traverses a large number of nodes and the planned path is not unhindered when it searches the path in the indoor large scene environment,an improved A*algorithm is proposed.By adopting the adaptive weighting method for the enlightening function,its proportion in the cost function is changed.Compared with the traditional A*algorithm,the number of nodes traversed by the improved algorithm is reduced by 50.4%.The method of smoothing the planned motion path by using Bessel curve is proposed to reduce the turning angle of the path.Compared with the traditional A*algorithm,the optimized path reduces the turning point and meets the movement requirements of the tracked mobile robot.Finally,the autonomous navigation experiment of the tracked mobile robot is implemented,and the mapping experiment of SLAM algorithm is implemented in the indoor scene.The experimental results show that the average error of the traditional Cartographer algorithm for the description of the feature object is 2.83%,while the average error of the improved Cartographer algorithm is reduced to 1.43%.The path planning experiment and dynamic obstacle avoidance experiment of mobile robot are designed,and it is verified that under the integrated navigation mode of global path planning and local path planning,mobile robot can safely avoid obstacles,successfully reach the target point,and meet the autonomous navigation requirements of mobile robot.