Research On Autonomous Navigation Technology Of Combine Harvester Based On Beidou Positioning

In order to improve the intelligence level of agricultural machinery and accelerate the realization of fully unmanned harvesting operation,this paper takes the wheeled combine harvester as a platform,builds an autonomous navigation control system around two major aspects of autonomous harvesting operation: straight-line harvesting and field steering,designs a humanoid driving model controller and adaptive pure tracking curve navigation,and conducts simulation verification,road functional tests and field tests on the navigation controller.The main research contents and conclusions are as follows.(1)An autonomous navigation system for combine harvester based on Bei Dou satellite positioning was built.The hardware system consists of a sensing module,a control module and an actuation module.The perception module consists of angle sensor and Beidou RTK dual antenna positioning module,which is responsible for receiving rear wheel rotation angle and position information;the control module consists of industrial control computer and NI data acquisition card,which is responsible for calculating and sending navigation control volume;the execution module is responsible for receiving steering signal and completing steering control.The navigation software is written in C/C++ language,which can realize the functions of controlling the rear wheel steering,path planning and tracking,human-machine interaction,data acquisition and saving,etc.Around the main thread harvester target corner calculation and control signal sending with multi-thread technology to realize the parallel processing of harvester path planning,real-time data saving and automatic navigation control,etc.(2)The autonomous operation navigation system of the combine harvester is designed.The requirements for autonomous operation navigation are clarified,the path of set row operation is planned,and the autonomous operation navigation system is designed,which includes a humanoid driving model navigation controller and an adaptive pure tracking curve navigation controller.The anthropomorphic driving model navigation controller designed an offline BP neural network by collecting artificial driving data,and obtained an anthropomorphic driving model by using the neural network for training.The collected artificial driving data without training was verified by the neural network,and the results showed that the anthropomorphic driving model has good predictiveness;the Simulink tracking simulation model was constructed and the simulation test was completed,and the results showed that The anthropomorphic driving model control has the characteristics of fast convergence and small overshoot.The adaptive pure tracking curve navigation controller is based on the pure tracking principle,the curve path tracking error model is established,and the adaptive pure tracking model is obtained by using the weighted optimization method to determine the relationship between each factor and the forward-looking distance.In order to ensure the stability of navigation control when the two controllers are switched,the threshold value for the switching of the two controllers is designed,and the optimal switching state of the two controllers is determined.(3)Functional tests on the road surface and field tests of the autonomous operation navigation system of the combine harvester were carried out.The results of the road curve navigation test showed that the average absolute deviation was 0.034 m and the maximum tracking deviation was 0.058 m.The results of the road straight-line navigation anti-interference test showed that the harvester could complete the adjustment within 1.5 s and maintain high-precision straight-line tracking when it was disturbed by human.The test results of field set autonomous operation navigation show that when the harvester speed is set at 0.65 m/s,the maximum position deviation of the straight-line harvesting part is less than 0.0914 m,the average absolute error is less than0.0419 m,the average absolute deviation of field head turning curve navigation is less than 0.0537 m,and the overshoot amount after turning is less than 0.0775 m.It can complete the autonomous operation in the area to be harvested and meet the field harvesting It can complete the autonomous operation in the area to be harvested and meet the requirements of field harvesting.