Research On Path Tracking And Steering Control Method Of Harvester Based On Kinematic Characteristics

The intelligent rice and wheat harvester driver assistance system has the advantages of high precision,autonomy and high efficiency.The driver assistance system is of great significance to the development of smart agriculture and maintaining food security.The harvester has the characteristics of large size,large inertia,time-varying load,and rear wheel steering.These characteristics bring challenges to the development and application of path tracking and steering control systems.Aiming at the problems of poor adaptive ability,system delay,susceptibility to interference in the path tracking and steering control system of the rice and wheat combine harvester,the kinematic characteristics of harvester are studied.The path tracking controller with heading adaptive calibration,the pure pursuit predictive path tracking controller,the steering controller with mass adjustment factor and the sliding mode controller based on a new reaching law are proposed,realizing the path correction and assisted driving of the intelligent rice and wheat harvester.The main contributions of this paper are as follows:(1)Aiming at the constant deviation of harvester driver assistance system caused by the calibration errors of heading and median angle,a path tracking controller with heading adaptive calibration is proposed.The error models of heading and median angle are established.The average lateral distance error is calculated by the self-adjusting rolling time window.The incremental Proportional-Integral-Derivative(PID)controller optimized by the improved Particle Swarm Optimization(PSO)algorithm is used to compensate the heading and median angle calibration errors,and restrain the constant deviation.Aiming at the problem of excessively fast adjustment of inertia weight of PSO,a new inertia weight adjustment strategy combining the number of iterations and the feedback function is proposed.The new strategy can avoid the algorithm from falling into the local optimum.The simulations demonstrate that the system can adaptively compensate the heading and median angle calibration errors under the conditions of different amplitudes and different directions.The constant deviation of path tracking controller can be eliminated and the effectiveness of path tracking controller with heading adaptive calibration is verified.(2)Aiming at the problem of curve path tracking control delay caused by the large inertia of harvester,a pure pursuit predictive path tracking controller is proposed.The harvester kinematic prediction model is established based on the kinematics characteristics.According to the harvester’s speed,heading,steering angle and control signal,the controller predicts the position and attitude of the harvester at the next moment,and corrects the control variable at the current moment.The prediction time adjustment criterion is proposed to optimize the correction coefficient to improve the real-time performance and tracking accuracy of harvester on the curved path.The simulations show that according to the predicted system state,the controller can correct the control variable,which can improve the adaptability and real-time performance of the driver assistance system to different curved paths.(3)Aiming at the problem of steering control precision reduction caused by soil compaction difference and the time-varying load of harvester,a steering controller based on the mass adjustment factor is proposed.The constant adjustment factor is used to characterize the effect of soil compaction difference,and the time-varying adjustment factor is used to represent the influence of load change of the harvester.The steering controller parameters are optimized by the mass adjustment factor.The simulations demonstrate that under the conditions of soil compaction difference and the time-varying load,the controller can adjust the control proportional coefficient in real time to mprove the steering control accuracy of the harvester,which proves the effectiveness of the steering controller based on the mass adjustment factor.(4)Aiming at the problem of control accuracy reduction caused by the disturbance of farmland unevenness,a new sliding mode steering controller is proposed.A new reaching law is proposed,and eliminates the chattering phenomenon by the exponential function.The control saturation problem caused by the large initial error is avoided by the initial value of the switching function.A disturbance observer is constructed for the real-time observation of disturbance signals,which makes the controller get rid of the dependence on the prior knowledge of the disturbance.The simulations show that the new reaching law sliding mode steering controller has a good performance on initial errors and impact interference errors.(5)The driver assistance system of intelligent rice and wheat harvester is developed,and the harvesting field experiments are carried out.The heading is calibrated adaptively by the path tracking controller with heading adaptive calibration,and the heading error is calculated according to the corrected heading.The desired steering angle is calculated by the pure pursuit predictive path tracking controller.The new sliding mode steering controller based on the mass adjustment factor controls the wheels to the desired steering angle,and the assisiting driving is achieved.Field experiments demonstrate that the path tracking controller with heading adaptive calibration can effectively restrain the constant deviation error,and the path tracking standard error is 1.82 cm.Under external interference conditions,the path tracking standard error of the new reaching law steering controller is 2.81 cm.The linear path tracking standard error of the harvester path tracking and steering control system is 1.72 cm,and the curve path tracking standard error is 3.84 cm,which has passed the inspection of national agricultural machinery quality supervision and inspection center.