Lidar-Based Positioning Algorithm And Module Realization

In recent years,artificial intelligence technology has developed rapidly and has been widely used in various advanced technology fields.Among them,in the field of unmanned systems,the development is most significant,with various types of unmanned aerial vehicles,unmanned ships,and unmanned boats successively appearing.The intelligence embodied by the unmanned system needs to complete tasks such as understanding the environment,determining its own posture,and making decisions and actions.This requires the deployment of various sensors on the unmanned system and the completion of data collection and processing.Among them,the solution of the positioning problem is the prerequisite for the unmanned system to become intelligent.In this paper,under the environment of poor GPS signal,the positioning algorithm of lidar is studied and the positioning module is designed,including the modeling of lidar environment,the research and realization of three kinds of laser positioning algorithms,and the car experimental platform based on lidar positioning module Build.The main tasks of this paper are:(1)Analyze various types of lidar and study the positioning model based on lidar.In view of the comparison of the working mechanism,applicable environment and performance of various types of lidar,Rplidar A2 lidar is selected as the research object and its working principle is analyzed;the positioning model based on lidar involves positioning modeling and map representation,and positioning modeling uses multiple The observation points are known to determine the pose of the next lidar,and the map model uses a grid map that uses a probability value to determine whether the grid is occupied by obstacles.(2)Research on the Lidar-based real-time localization and mapping(Simultaneous Localization And Mapping,SLAM)algorithm,and conduct simulation experiments based on the Gazebo simulator.SLAM algorithms can be divided into particle filter-based algorithms and graph-based optimization algorithms.Modeling and analysis of these two types of algorithms;research on particle filter-based Gmapping algorithm,graph-based optimization Hector SLAM algorithm and Cartographer algorithm,and implement these algorithms Step analysis,and finally through the indoor positioning mapping experiment on these algorithms based on the gazebo simulator.(3)Researched the trolley platform based on the lidar positioning module;including the structure design of the positioning module based on the lidar and the structure design of the trolley chassis control system based on the Open RE Board main controller;then introduced the positioning module and the underlying control based on HFLink System communication method;then design the trolley platform software;finally,analyze and simulate the path planning experiment of positioning module on autonomous navigation,including global and local path planning,global path planning uses A＿star algorithm,and local path planning uses dynamic window algorithm.(4)The design of positioning experiment scheme based on positioning module,the analysis of experimental results and the experimental analysis based on positioning navigation.Four positioning schemes are formulated according to three types of positioning algorithms: Gmapping experiment(odometer + IMU +positioning module),Hector handheld experiment(positioning module),Hector plane movement(positioning module),Cartographer experiment(IMU + positioning module),and in these four Ten sampling points were calibrated in one experiment,and the positioning accuracy of these four sets of data was analyzed;the realization steps of positioning and navigation were given and experimental analysis was carried out.