Research On Navigation Technology Of Indoor Mobile Robot Based On SLAM

Navigating autonomously of mobile robots is a symbol of the development direction of high and new technology,which is in the forefront of the field of scientific research.And,navigating autonomously based on Simultaneous Localization and Mapping(SLAM)is a problem that has attracted a great deal of interest in mobile robotics.In order to make the indoor mobile robot more autonomous and intelligent,this topic will study the indoor navigating autonomously based on SLAM,and further study the three problems of SLAM,localization and path planning.Due to its modular design,it has many different applications such as service robots and transport robot.The completion of this thesis will promote the development of autonomous navigation in China.Firstly,in view of the shortcomings of the traditional filter-based SLAM algorithm,a SLAM algorithm based on graph optimization is studied in this thesis,and utilizes it to estimate the pose of the robot and the location of the features.Key steps in SLAM such as data association,loopback detection,and backend optimization are analyzed.In this paper,an active SLAM technology is designed,and the active exploration method based on the boundary is used to combine the SLAM with the path planning,so that the robot can build the map of the unknown environment autonomously,so that the robot can be truly autonomous.In addition,the basic Monte Carlo localization algorithm and adaptive Monte Carlo localization algorithm are studied.And the convergence process of localization algorithm is analyzed through simulation experiments.The global planner algorithm A* is studied and compared with other algorithms through simulation experiments to verify its superiority.A novel local path planner algorithm is applied to achieve real-time trajectory optimization,and simulation experiments are performed to verify its effectiveness.In order to navigate safely in indoor environment,the thesis presents a method for achieving 3D obstacle avoidance in indoor complex environments,which is based on the three dimensional point cloud.Finally,all the algorithms presented in this thesis have been tested and verified in real world experiments.Active SLAM experiments prove the feasibility of active SLAM technology.By comparing and analyzing the maps obtained by the two SLAM methods,the superiority of the SLAM algorithm used in this paper is explained.A comparison of the key dimensions between the map and the actual environment verifies the accuracy of the map.Autonomous navigation experiments verify the performance of localization algorithm and path planning algorithm,and the navigation accuracy is analyzed.Through the obstacle avoidance experiment,the navigation system's 3D obstacle avoidance ability and dynamic obstacle avoidance ability are tested.It is proved that the real-time and robust of the path planning algorithm is good.Through the research on SLAM,localization,and path planning,this thesis has finally obtained a stable navigation system with high autonomous capabilities and high navigation accuracy.