Research On Automatic Driving System Of Crawler Tractor Based On Double-laser Positioning

In the background of agricultural modernization, Agricultural machinery development is in the direction of precise and intelligent. This paper summarizes the research achievements of automatic driving technology in the domestic and abroad and take the Nongfu NF502 crawler tractor as a platform to study the automatic driving technology of crawler tractor based on double laser positioning system.At present, agricultural machinery automatic driving technology research focus on wheeled tractor and most use GPS as a navigation system, but crawler tractor automatic driving technology research is relatively few. In this paper, the kinematics model of wheeled tractor and crawler tractor are introduced. Wheeled tractor makes turns by turning angle with front wheel while crawler tractor makes turns by both sides caterpillar to different sports modes. This paper analyzes the main parameters in the process of movement of crawler tractor, and the kinematics model is established. Automatic driving algorithm combines cross coupling algorithm with fuzzy algorithm to realize scheduled path tracking.Crawler tractor automatic driving system consists of two parts, control system and the executing agency. Control system includes MCU, double source laser positioning system, sensor, the peripheral circuit, etc. The executing agency include HST speed control system, ignition/choke control system, hydraulic steering system, etc.HST speed control system uses adaptive fuzzy PID algorithm to speed forward and backward motion control and steering control is realized by using fuzzy adaptive algorithm.The calibration experiment and error compensation electronic for compass, gyroscope, HST electric putter better meet the tractor working environment and improve the control precision. To verify the validity of the path tracking algorithm, the path tracking experiment is tested. In the test, the speed of the tractor is 1m/s and the driving distance is 150m, and the running route of the tractor is recorded. The test results show that the average tracking error between the actual running route and the scheduled route is less than 20cm, and the highest accuracy is 5cm.The test result shows that the algorithm can achieve a better control of the tractor, the control effect is efficient.