Research On Map Construction And Positioning Of AGV Based On Laser Natural Navigation

Since "made in China 2025”promulgated,intelligent manufacturing is defined as the main direction of manufacturing in China,using intelligent manufacturing technology for manufacturing process automation upgrading has become main improvement direction of traditional manufacturing industry.China's industrial automation level has been significantly improved,including Automated guided vehicle(Automated guided vehicle,AGV)as a new type of intelligent,Automated transport device,it has got rapid development,the AGV in manufacturing applications shows the development trend of industrialization.At present,the mainstream navigation method in AGV application is laser navigation.Based on this background,this subject has carried out the research of natural navigation AGV map construction and positioning algorithm based on laser sensor.In this paper,a laser natural navigation AGV driven by a single steering wheel is established(hereafter referred to as "laser navigation AGV)nature of the perception of the kinematics model,laser sensor model,and then defines the natural coordinate system model of AGV navigation,laser and laser between natural navigation AGV local coordinate system and the global coordinate system of the coordinate system conversion,thus achieved the lidar data point to the global coordinate system conversion,complete the reading of the lidar data frame,in this paper,the SLAM problem in adjacent time point scanning laser scanning data matching process by normal distribution transform algorithm to achieve,and the corresponding relationship of don't need to establish clearly,and Newton iterative algorithm is used to solve the optimal transformation parameters,The environmental map is described by the combining of NDT map and occupied grid map.Based on the map has been built,analyzed the principle and implementation steps of the monte carlo localization based on grid,and the current mainstream of the realization of the MCL is based on grid AMCL monte carlo localization(adaptive)experiment test,and error analysis of the reasons.In this paper,NDT algorithm is introduced into MCL algorithm to achieve global fast location,and use it to subsequent experiment.Take the built laser natural navigation AGV as the mobile platform,through the way of manual and automatic control of its movement,set up the lab map and real environment information at compare,Verify the accuracy of the map,and by comparing the selection of industrial application based on the location information obtained from natural navigation and localization algorithm,the paper proposes the position of the information,respectively for the repeated positioning accuracy and absolute positioning accuracy testing,verified the algorithm in the above two kinds of experimental test results meet the demands of industrial application accuracy.