Research On Path Planning And Path Tracking Methods Of Intelligent Harvester

Autonomous path planning and path tracking are the key technologies of the intelligent harvester,which directly affect the accuracy of the autonomous operation of the harvester.Focusing on the problems of low effectiveness and low versatility of the autonomous path planning,and poor convergence in the initial path tracking stage of the harvester,research on the path planning and path tracking methods of the intelligent harvester is carried out.Full coverage path planning method based on irregularly shaped farmland and path tracking method based on improved double tangent circle are proposed.The main research contents are as follows:(1)Modeling of working environment and path planning of headland turnBy collecting the boundary points of the irregular quadrilateral farmland,the contour model of the autonomous operation area is constructed.The optimal operation direction is selected based on the least number of turns and the smallest turning distance,and the construction of operation line model is preliminarily completed.The operation line model is perfected by estimating the grain output of each operation line,and then the modeling of the operation environment can be completed with the outline of the work area.The shortcomings of traditional turning methods in the irregular quadrilateral boundary are analyzed,and an improved U-shaped turning path is designed to reduce the turning area of the headland.(2)Full coverage path planning under multi-objective constraintsThe traditional full-coverage path form is studied.The influence of different grain unloading position distribution,the degree of compaction of the harvester on the land and the capacity of the harvester on the full-coverage path are analyzed.A method for solving the traversal sequence of operation rows under multi-objective constraints is proposed.According to five kinds of unloading location distribution,the distance model between the operation rows and the unloading positions is constructed.taking full-load distance constraints and capacity constraints as conditions,a corresponding solution model is constructed.According to the relationship between the minimum turning radius of the harvester and the width of the operation row,the traditional shuttle method and the improved ant colony algorithm are used to solve the traversal sequence under the constraint condition of full-load distance and capacity.The method can reduce the degree of crushing on the ground,and improve the adaptability of path planning algorithm to the different unloading location distributions.(3)Design of double tangent circle optimization algorithmAiming at the selection of preview points for large-angle steering control method,the basic principles of the double tangent circle path planning algorithm are studied.An automatic optimization global planning algorithm is designed to solve the optimal target position.By considering the steering delay of the harvester and the actual tracking error,a local planning algorithm is proposed to adjust the tangent points,and preview points are provided for the largeangle steering control method.The method can realize fast tracking when tracking deviation is large.(4)Design of hybrid path tracking method for intelligent harvesterThe geometric principle of pure pursuit algorithm is studied,and pure pursuit algorithm based on the kinematics model of the harvester is designed by combining the characteristics of the harvester.A hybrid path tracking method that combines large-angle steering control and pure pursuit is proposed by analyzing the problem of poor path tracking convergence when the initial error is large and exploring the error mechanism of the path tracking.Track the path according to the lateral deviation and the heading deviation.If the tracking deviation is large,the large-angle steering control method is selected to achieve fast tracking.If the tracking deviation is small,the pure pursuit algorithm is used to achieve high-precision path tracking.