Headland Turning For Agricultural Robot Vehicles Based On Laser Radar Sensor

With our country industry to keep pace of times and the continuous stepped-up in-depth development of the industrialization and urbanization and the improvement of people's living standards,the rural young adults labor go to the cities and integrate into the city life,which make the rural population loss become worse,and cultivation of land urgently need to develop agricultural mechanization.As a part of the agriculture,the development of orchard agriculture need to promote automation and intellectualization of agriculture robot vehicle,which would improve the orchard production efficiency in pruning,spraying and harvesting,as well as reduce labor costs.At present,most studies focused on the agricultural robot vehicle autonomous navigation between two row fruit,less studies carried out headland turning path planning in the orchard,but it is crucial important to enter into the next line for agricultural robot vehicle autonomous path planning and positioning correctly.Based on the differential steering of agricultural robot vehicle,this paper studied on headland turning path planning with the environment of the density of Peach trees orchard.It is also a supplement of the content of orchard vehicle autonomous navigation.The main contents of this topic are:1.The paper studied detailed the orchard agriculture robot vehicle platform and the main equipment including the PWM pulse width signal isolation transmitter,the characteristics and distance measuring principle of the laser radar sensor,and the communication with PC,driver module,upper machine and lower machine,circuit connection of hardware among agricultural robot vehicle.2.This paper studied LMS291-S05 laser radar sensor parameters configuration testing and the method,the baud rate configuration commands,scanning angle,scanning angle resolution configuration commands,data transfer instructions and data output the syntax of the format,etc;it also studied calibration principle of laser radar sensor and chose metope as a calibration template,set the minimum resolution scanning angle to extract fruit trees location information and make the agricultural robot vehicles walk independently between fruit tree row.3.Kinematic model of agricultural robot vehicle was established.The agricultural robot vehicle reached headland which was precisely determined,and headland width was measured.The paper studied path fitting of agricultural robot vehicles which walked between fruit trees and adopted fuzzy control way to walk,and studied the working principle and strategy choice of agricultural robot vehicles walk straight and headland turning.4.Headland turning path was analyzed by simulation according to kinematics equation of agriculture robot vehicle.Generated turning path of U-turn under the sufficient width and K-turn under insufficient width of agricultural robot vehicle were analyzed respectively.It also studied the basic principle of analytic hierarchy process,and evaluated each generated simulation path,determined the weight of each path under the respective conditions,it is concluded that the optimal path by comparing the weight.5.Agricultural robot vehicles walk on the fruit trees under the condition of whether there is offset in the initial starting position,the stability of the fuzzy control algorithm is validated.Under the condition of insufficient and sufficient headland width,the experiment of U-turn and K-tum were carried out respectively and the experimental results were analyzed.The average minimum direction angle and lateral error are 15.6°,13.7cm and 13.7°,13.9cm for U-turn and K-turn respectively.