Research On Autonomous System Of Agricultural Machinery In The Field For Smart Farm

As a key technologies in the field of autonomous farm machinery field systems,the autonomous farm machinery full-coverage path automatic planning technology for smart farms plays an important role in improving the efficiency and quality of farm machinery operations such as tillage,fertilization and harvesting,and reducing the energy cost of operations.Precise positioning is the core basis for realizing autonomous farm machinery,and its positioning accuracy and stability will directly affect the reliability and safety of farm machinery driving;while path planning is the key to realize autonomous farm machinery operation,and the superiority of algorithm will directly affect the quality and efficiency of farm machinery operation.This thesis takes autonomous farm machinery as the research object,and aims to improve the positioning accuracy and stability of autonomous farm machinery as well as the quality and practicality of path planning,whose main research contents include:(1)studying the positioning navigation theory and positioning system based on RTK technology,determining the use of positioning navigation device based on the fusion of RTK and heading projection to collect positioning information,and deriving the principle and conversion equations between WGS-84 coordinate system,geodetic plane coordinate system and carrier coordinate system to provide position coordinates for autonomous farm machinery path planning,using Kalman filter to noise reduction processing of the collected positioning data to The Kalman filter is used to reduce the noise of the collected positioning data to achieve continuous and stable precise positioning,and the static test proves that the real-time position with centimeter-level accuracy can be continuously and stably output for the autonomous farm machinery.(2)The raster environment map of autonomous farming machine operation is established by using farmland boundary coordinates,and full-coverage path planning is carried out based on improved genetic algorithm.The simulation experimental results show that the method can effectively guide autonomous farm machines to achieve efficient full-coverage operations under complex environmental maps.(3)For the existence of the minimum turning radius of autonomous farm machinery,we analyze the Ackermann kinematic model and study the smooth path planning algorithm satisfying the incompleteness constraint,and decide to adopt the multi-objective cost function based on the path smoother and introduce the Reeds-Shepp curve frontal heuristic function to optimize Hybrid A~* as the planning algorithm for local path smoothing of autonomous farm machinery,and the simulation experiments The results show that the planned paths are smoother and conform to the actual operation scenarios of autonomous farm machines.(4)The experimental vehicle platform of the autonomous agricultural machine is built and the functional requirements and overall design framework of the platform are analyzed,and the experiments are conducted in three aspects of algorithm convergence,algorithm feasibility and path optimality,and finally the comprehensive energy consumption is verified and analyzed in the experimental vehicle.The results show that the planned full-coverage paths are on average 38.54% and 35.00% lower in terms of repeated operation area and number of turnarounds,respectively,and only the number of turns is increased by 13.76%,but the comprehensive energy consumption is reduced by 7.82%,which shows the superiority of the algorithm compared with other algorithms.This thesis investigates three aspects of autonomous agricultural machines:precise positioning,full-coverage path planning algorithm and local path re-optimization considering path smoothing,which can provide path planning guidance for autonomous agricultural machines with practical significance.