Path Planning And Path Tracking Of Intelligent Agricultural Vehicle Based On Multi-Source Information Fusion

The self-driving operation of intelligent agricultural vehicle can reduce the dependence of agricultural production on the labor force and the working intensity of agricultural employees in the busy seasons.What’s more,it is the best way to improve the efficiency of agricultural production.The recent researches are mainly aimed at the ability of agricultural vehicle to track the reference trajectory.Which lack about the effect of the actual operating environment on the self-driving of agricultural vehicle.The farmland traversal always using the traditional landhead turning method,and don’t fully consider the conditions of the landhead turning to the boundary space,on the other hand the influence of the soft pavement on the effect of the tire usually neglected.The researches are still at the level of technology research,and the practical operation of the algorithm verified is lacked.The main research contents of are demostrated as follows:(1)An intelligent agricultural vehicle test platform was established,and the system composition of the platform was designed,according to the basic requirements of lateral motion control,design environmental perception system,multi-source information fusion system,path planning decision-making system and lateral control system.(2)The actual environment of the standard orchard was investigated.According to the specific situation of the internal and boundary information of the standard orchard field,the traversing mode and the way of the landhead turning were synthetically analyzed.The vehicle routing problem(VRP)mathematical model of the current logistics industry was adopted.The branch-and-bound method is used to solve the VRP problem,which realizes the global path planning of the standard orchard.(3)Considering the soft characteristics of the driving road,the interaction between the tire and the soft pavement is analyzed,and the experience model is used to calculate the tire force,which provides the basis for the lateral control of the intelligent agricultural vehicle.According to the requirements of the lateral control,the model predictive control(MPC)is used to design the path tracking controller with lateral dynamic model and a path tracking error dynamics model,and the Simulink-Carsim joint simulation platform is built to perform the simulation experiment,and the "S" path and the "8" shape path are taken into consideration.The experimental results show that the designed controller can realize the landhead turning of the path of large curvature,and it has good operation stability.(4)In order to verify the practical maneuverability of the algorithm,the subject takes the Huanghaijinma 404 C tractor as a platform,and the automatic steering system is reformed.The field test is carried out on the cement pavement and the soft grassland.The experimental results show that the calculated speed of the designed controller is high in real-time,and it can effectively control the intelligent agricultural vehicle to self drive on the desired path.