Research On Map Construction And Navigation Method Of Mobile Robot Based On Two-dimensional Code

As an important member of robots,mobile robots have been widely used in security,housekeeping services,logistics and other fields.Path planning and SLAM(Simultaneous Localization and Mapping)technology are the two key technologies in the research direction of mobile robots.They are the basis for completing other complex tasks.Therefore,the research of map building and path planning in large indoor environment has important practical value and theoretical significance.The thesis firstly analyzes the current research status of mobile robot mainstream navigation methods,SLAM and path planning technology,and determines the research direction of the thesis.Then we design a mobile robot with two-dimensional code based on ROS operating system.The robot hardware system can be divided into upper,middle and lower layers according to the classic embedded model.In the hardware system,mathematical models of the chassis,odometer,laser observation sensor and monocular camera are built according to functional modules,and the robot's motion model and sensor observation model are obtained.Aiming at the problem that the traditional laser SLAM method will produce large cumulative errors in large-scale environment,which leads to the failure of map construction,an improved SLAM scheme is proposed which integrates monocular camera.The monocular camera can obtain rich feature information from the surrounding environment to accurately determine the actual closed-loop point.In the loop detection process,the laser scanning data is merged to establish strong constraints between the nodes of the pose map,so as improve the mapping accuracy and making up for the shortcomings of the less information collected by the lidar.But the vision scheme generally has the shortcoming of low real-time performance,so this scheme uses two-dimensional code as artificial landmark to assist laser data loop detection.Based on the above,the plan selects the continuous L-M back-end optimization method to establish a map consistent with the environment.In order to solve the navigation problem of the mobile robot in the unknown environment,part of the target point of the known or known map.Based on the traditional AFSA algorithm,the thesis proposes an AFSA algorithm that dynamically adjusts the field of view and congestion factor.At the same time,the weight factor is introduced to improve the social behavior of artificial fish swarm,so as to avoid the traditional algorithm's fixed field of vision and step size,which is easy to fall into local optimality,to improve the search accuracy and computing efficiency of the algorithm.At the end of the thesis,the two-dimensional code mobile robot developed by the laboratory is used as the carrier,carrying the traditional laser map optimization SLAM and the improved method,taking the two-dimensional code as the target navigation point to carry out the map construction and path planning work in the laboratory and teaching building.The experimental results show that the improved method has high stability and reliability.