Research Of AGV Intelligent Navigation System Based On Laser SLAM

In recent years,with the deep integration of industrial technology and automation technology,the research and application of intelligent industrial robot has become an upsurge in the field of robot exploration.As one of the representatives of industrial robots,Automatic Guided Vehicles(AGV)is widely used in manufacturing and transportation industries.However,the localization and navigation mode of AGV in China has some problems now,such as single navigation path and unable to realize intelligent avoidance,which can not meet the development requirements of intelligent and flexible modern factories.Based on this,this paper proposes an intelligent navigation system based on laser slam,which takes lidar as the core sensor,senses the environment by using simultaneous localization and mapping technology,constructs a two-dimensional map,then uses the constructed map for path planning to achieve the goal of intelligent navigation.The main work of this subject is as follows:1.According to the research needs of this subject,a four-wheel all drive intelligent moving platform is built independently,and the hardware selection is analyzed.The operating system of ROS is deeply studied,and the intelligent navigation scheme of the system is designed.2.The moving control and data processing of intelligent moving platform are modeled.The four-wheel differential moving model of the moving platform,the coordinate transformation model and the lidar optical measurement model.3.Aiming at the problem of location and mapping of indoor unknown environment,the technology of simultaneous location and mapping is used to construct two-dimensional map.Firstly,through comparative analysis,the probability grid map is selected as the map representation of this subject,and the construction principle is analyzed.Then,cartographer algorithm is used to construct map of environment,and its principle is analyzed theoretically.4.Aiming at the problem of navigation in known environment map,a navigation system based on Navigation framework is designed.The system mainly includes global path planning and local path planning.The global path planning uses A* algorithm to plan an optimal route in the map.The local path planning uses dynamic window method(DWA)to avoid dynamic obstacles in the environment in real time.Then the A* algorithm is optimized to eliminate redundant inflection points and improve the navigation efficiency.5.The experimental test of intelligent navigation system is completed.The simulation model of moving platform and simulation environment are built in ROS,and the simulation experiments are carried out Then the laser slam mapping and navigation experiment are carried out in the real environment by using the four-wheel all drive intelligent moving platform.The experiment proves the effectiveness of the algorithm of this subject.The function of remote display and control of robot state is realized by remote display and control software.