Research And Implementation Of Robot Autonomous Exploration Mapping System Based On Factor Diagram

Autonomous exploratory mapping system is one of the core technologies in Exploring Robot applications,which is used to build unknown environment maps and provide environmental information for other applications.At present,some progress has been made in the research of autonomous exploration mapping system,but it still faces serious challenges in complex scenarios.On the one hand,complex scene reconstruction requires higher positioning accuracy of the robot.On the other hand,complex scenarios require more efficient and robust exploratory path planning algorithms.In addition,the autonomous exploratory mapping system should be able to handle exploratory tasks with different sets of high-altitude scenes to achieve three-dimensional scene reconstruction and path planning.To solve the above problems,a selfexploring mapping system for mobile robot based on laser inertial navigation is designed,and the mapping problem for simultaneous positioning and the path planning algorithm are optimized.The main content and the work done are as follows:A laser inertial navigation tight coupling simultaneous positioning and mapping system based on factor diagram is designed to achieve decimeter level robot positioning and point cloud map construction by fusing lidar and inertial measurement unit information.The system uses a factor diagram as a back-end optimizer to fuse lidar and inertial measurement unit data.A curvature-based point cloud matching algorithm is designed to improve the robustness of the inter-frame milemeter.A loop detection module based on SCAN-CONTEXT is introduced to overcome the effect of accumulated error.An autonomous exploratory path planning algorithm based on hierarchical thought is designed.The method divides the exploration path into local and global exploration paths.Local search paths are generated based on candidate observers and global search paths.After the optimal observation points are obtained,the visiting order of the observation points is calculated by solving the traveling salesman problem,and then the optimal search paths are solved by the track smoothing algorithm based on the minimum snap.The global search path calculation contains the global optimal paths for all areas not explored,providing additional observer information for local search paths.An experimental platform for autonomous exploration and mapping based on wheeled differential mobile robot is designed and implemented.Sensor selection and hardware and software adaption are completed.The modules of obstacle detection,local path planning and path follower based on lidar are implemented,and an experimental platform for mobile robot based on ROS is built.A real vehicle experiment is carried out on a mobile robot autonomous exploration mapping system based on laser inertial navigation.The experimental results show that the system has high positioning accuracy and mapping accuracy in complex scenes,and can efficiently complete the task of autonomous exploration mapping in unknown scenes.