Research On Autonomous Navigation Technology Of Mobile Robot Based On Laser SLAM

With the continuous improvement of the level of science and technology and the rapid development of social productivity,robots and artificial intelligence technology have received more and more attention and application in daily production and life,research and breakthroughs in the field of robot are also becoming strong driving force to promote China’s industrial upgrading and replacement of growth.As an important branch of robot technology,mobile robot has become a hot direction and frontier field of current robot and artificial intelligence technology research due to its interdisciplinary research characteristics and multidimensional application scenarios.In order to solve the problem of intelligent and autonomous driving of mobile robot,this paper studies the autonomous navigation technology based on laser SLAM,and realizes the map construction,autonomous positioning,path planning and real-time obstacle avoidance of mobile robot in unknown scenes.The main research contents of this thesis are as follows:Firstly,in order to solve the problem of map construction in unknown environment,the principles and models of lidar and occupancy grid map are discussed in detail,and the theoretical basis and interaction mechanism of each module of the graph optimization method are mainly studied.On this basis,an improved graph optimization method based on adaptive loop matching strategy is proposed.This method adaptively adjusts the loop matching search step size by using the geometric loop degree of the map construction route,and combines the validity judgment to improve the accuracy and reliability of loop detection.The simulation mapping experiments of the algorithm are designed on the open source map data set to verify the back-end optimization effect of the graph optimization method and the overall improvement of the global mapping accuracy of the improved graph optimization method.Secondly,in order to solve the problem of autonomous robot’s path planning in known maps,the model of adaptive Monte Carlo positioning algorithm and the path planning environment model based on cost map are studied,so as to provide the basis of position and environment information for path planning.Through the theoretical analysis and simulation comparison of the relevant algorithms of global path planning,the superiority of A*algorithm compared with other path planning methods is verified.And by optimizing the travel cost prediction function and path generation method of the A*algorithm,the global planning efficiency and path generation effect of the algorithm are improved.In this thesis,the speed sampling mechanism and trajectory evaluation method of DWA local path planning algorithm are studied in detail,and the improved effect of optimized A*algorithm and the performance of local planning with real-time obstacle avoidance are further verified through the virtual scene simulation experiment on the gazebo platform.Finally,in order to verify the operation effect of the relevant algorithms from a practical point of view,this thesis completes the overall design of the mobile robot autonomous navigation experimental platform based on the hardware framework of ARTROBOT smart car and the software framework based on ROS system,and on this basis,tests the SLAM map construction,global path planning and local real-time obstacle avoidance in the indoor scene.By comparing and analyzing the results of SLAM and path planning experiments,the actual performance of each algorithm in this paper and the overall optimization effect of the related improved algorithms are verified,and the obstacle avoidance ability of the smart car under different dynamic obstacle road conditions is experimentally verified to ensure that the autonomous navigation of the mobile robot complies with the real-time requirements of realworld scenarios.