Adaptive Control Method For Tractor Auto-guidance System

The tractor automatic navigation system was a research hotspot in the field of intelligent agricultural equipment.It helps to promote production efficiency,improve the accuracy of operations,and meet the needs of large-scale land operations.This paper focuses on the adaptive control method of automatic navigation system to solve the problems of poor robustness of navigation controller and high failure rate of front wheel angle sensor.The specific research contents are as follows:1.Adaptive sliding mode control method for navigation end-effector.To solve the control problem of the navigation end actuator under the condition of nonlinear interference such as mechanical clearance,dead zone and friction,the mathematical model of the controlled objects was studied,a discrete sliding mode control algorithm was designed based on the exponential reaching law,an on-line interference compensator was designed to identify and compensate the uncertain interference,and then realized adaptive control of navigation end-effectors.The simulation results showed that this method solved the chattering problem of the traditional sliding mode controller,could effectively resist the external nonlinear interference,and improve the system robustness.2.Multi-sensor data fusion measurement method for front wheel angle.To solve the problem of control misalignment caused by high failure rate of the front wheel angle sensor,the dynamic steering characteristics of the tractor were analyzed,and the relationship between the tractor attitude information and the front wheel rotation angle was established.A new hybrid Kalman structure was designed for multi-sensor data fusion measurement of the front wheel rotation angle,based on the conventional Kalman filter algorithm and the robust weighted observation fusion Kalman filter algorithm.The test results showed that the algorithm automatically switched to the fusion measurement output when the front wheel rotation angle sensor failed.The maximum estimation error was 0.23° and the root mean square error(RMS)was 0.13°.It improved the ability of the navigation controller to adapt to external disturbances.3.Improved the pure pursuit algorithm.To solve the problem of poor self-adaptability in path tracking by traditional pure pursuit algorithm,an improved method was proposed based on the seeker optimization algorithm(SOA),so that the front view distance could be dynamically adjusted according to the fitness function.The comparison simulation results showed that the improved pure pursuit algorithm could improve the control system's self-adaptability and reduce the path tracking error when the vehicle speed changed dynamically.4.Real-time dynamic optimization of online trajectory planning method.To solve the problem of the long distance traveled by the previous path tracking algorithm and the poor control accuracy near the upline point,a real-time dynamic trajectory planning technique was used to improve the path following control method.Considering the constraints of vehicle kinematics,the optimal trajectory planning problem was transformed into the parameter optimization problem of B-spline control points.The parameters of the control points were optimized by the quantum genetic algorithm to obtain the trajectory that satisfies the requirements.The simulation results showed that the algorithm was suitable for a variety of operating conditions,and the convergence speed was fast,and could meet the real-time planning requirements.5.Position-direction dual closed-loop cascade sliding mode control method.To improve the self-adaptive ability of the path-tracking algorithm further,an internal-external ring-sliding-mode path tracking control system composed of a heading controller and a position controller was designed based on a hyperbolic tangent function.Simulation results showed that the proposed algorithm could adaptively adjust the speed and yaw rate of agricultural vehicle so that the vehicle could track the desired trajectory quickly and accurately.6.Design of automated navigation system and field trials.To demonstrate the correctness of the algorithm from the perspective of field trials,the software and hardware of the automatic navigation system were designed,and the LOVOL M1004 wheeled tractor and the NF-752 crawler tractor navigation test platform were developed.The results of path-tracking test showed that the maximum linear tracking error of the wheeled and crawler tractor navigation system was 3.2cm and 3.3cm respectively,and the RMS was 1.3cm and 1.8cm respectively;The maximum curve tracking error of the improved pure pursuit algorithm was 10.0 cm and RMS was 6.0 cm;The maximum curve tracking error of the double-loop sliding mode path tracking algorithm was 9.0cm,and RMS was 4.0cm.It could be seen from the experimental data that the adaptive control method proposed in this paper could meet the needs of field navigation operations.