Research On AGV Hybrid Navigation And Positioning System In The Context Of Factory-based Rice Seedling Cultivatio

With the continuous development of information technology,rice factory seedling cultivation comes into being to meet the convenience of farmers.Under the large-scale,intensive and industrial mode of seedling cultivation,the problems of rice plate transportation and production efficiency are gradually reflected.Due to the large volume of the rice plate itself,the weight of the rice plate after loading the seedlings and soil layer can reach 4～6kg.It takes a lot of time and manpower to transport these rice plates one by one from the seed plate production line to the germination darkroom,as well as the subsequent transfer process.Reducing manpower input and labor intensity has become an inevitable trend.The common transportation mode can be divided into rail and trackless navigation.Rail needs to lay track,the route is not flexible,and the idle time is long.Therefore,this paper chooses the trackless transportation mode to study.According to the current situation of automatic rice seedling tray transportation in Qianjin Farm Co.,LTD.,Heilongjiang Beidahuang Agricultural Reclamation Group Company,and the production environment of rice seedling plant entering the farm as the research background,an automatic guided vehicle(AGV)transportation system with laser radar inertia and vision mixed navigation and positioning was designed.The research contents included the following aspects:(1)Based on the analysis of the working environment and the environmental factors that may affect the transportation of rice plates in the advanced farm,a scheme of laser radar inertial and visual hybrid navigation and positioning was proposed.(2)According to the actual working environment,the system is designed in detail,including the construction of motion structure,the selection and design of control system components,the selection of required functional modules,etc.,to achieve the required functions and complete the whole AGV construction.(3)The coordinate system involved in the operation of AGV in the production house was defined and transformed.The Lidar inertial navigation system was used to realize the operation of AGV on the specified path,and the operational attitude of AGV was solved by integral.Aiming at the noise existing in the inertial navigation system,Kalman filter algorithm is used to correct the Angle deviation and distance deviation in the inertial navigation of laser radar.(4)A series of artificial beacons for AGV navigation and a beacon recognition algorithm based on convolutional neural network are designed to further improve the accuracy of navigation and positioning on the basis of laser radar inertial navigation,and to solve the problem that beacons are polluted and unable to read information due to factors such as large dust of materials in the production workshop of automatic planter in the use of AGV.Through tests and experiments,the AGV system which uses Lidar inertial navigation and visual navigation together has high reliability and stability,can effectively adapt to the complex transportation and production environment,and complete the work requirements of transporting rice plates to the germination chamber on time and on demand,which has certain practical significance for the research of intelligent transportation.