基于速度自适应的拖拉机自动导航控制方法

Method on automatic navigation control of tractor based on speed adaptation

针对速度因素对拖拉机自动导航系统稳定性的影响,提出了基于横向位置偏差和航向角偏差的双目标联合滑模控制方法,在建立两轮拖拉机-路径动力学模型和直线路径跟踪偏差模型的基础上,应用Matlab/Simulink进行整体系统仿真,验证了控制方法的可靠性;以雷沃TG1254拖拉机为载体搭建了自动导航控制系统田间试验平台,分别在定速和变速条件下,进行了拖拉机直线路径跟踪控制的田间试验;分析了不同速度条件下的动态跟踪控制效果,验证了设计的自动导航控制系统的稳定性和控制精度。试验结果表明:在拖拉机田间作业常见的定速直线行驶工况下,采用基于速度自适应的双目标联合滑模控制方法,拖拉机直线路径跟踪控制的横向位置偏差最大值为10.60 cm,平均绝对偏差在3.50 cm以内;航向角偏差最大值为3.87°,平均绝对偏差在1.70°以内;在进入稳态以后,前轮转向角最大摆动幅度为3°,摆动标准差为0.80°。结论表明,该文提出的基于速度自适应的拖拉机自动导航控制系统,能基本实现不同速度下的直线路径自动跟踪控制。

It has been widely accepted that the application of tractor automatic navigation technology plays a key role in promoting the development of precision agriculture.The following fact is the increase in tractor operation speed and improvement of working efficiency.Considering the diversity of terrain condition and environmental noises,however,the nonlinearity and instability of tractor automatic navigation control system might be magnified significantly along with higher tractor speeds.Conventional control methods have been found to be difficult to meet the requirements for tractor automatic tracking.To investigate the influence of tractor speed on system stability,the sliding mode method was proposed based on the integrated control algorithm of deviations of lateral position and course angle.Firstly,deviation model of tractor linear path tracking was established based on the two-wheel tractor dynamic model with the introduction of velocity.The motion of tractor was simplified as the description of motion in two-dimensional plane,and only3degrees of freedom,i.e.the longitudinal and transverse direction and yaw of the tractor,were considered and some simplifications were made in the process of modeling.The visibility distance was set along the tractor forwarding direction.According to the kinematic relationship,the deviations of tractor lateral position and course angle at the visibility distance were obtained.In the actual path-tracking control process,it was difficult to achieve the ideal control effect at the same time.Therefore,the mixing coefficient was introduced into the controller to adjust the mixing degree of the2control strategies.Adaptive control of different deviations and speeds could be achieved when the coefficient changed between0and1.When it approached0,the control system tended to complete lateral position control;while the strategy of course angle deviation control was preferred with its value approaching1.The reliability of the proposed approach was verified by simulation using MATLAB/Simulink.The simulation results show that with the disturbance of tractor speed,rapid and stable tracking could be achieved for low speed conditions,too.Furthermore,an automatic navigation test platform was built on a Foton Lovol TG1254tractor.For tractor operation conditions at constant and variable speeds,a series of field experiments were conducted for the straight-path tracking control.By analyzing the effect of dynamic tracking control at various speeds using the mean absolute deviation,the maximum deviation and the standard deviation,the performance of the proposed automatic navigation control system was verified.Experimental results showed that in the tractor path-tracking process,the maximum deviation of tractor lateral position was10.60cm;the mean absolute deviation of lateral position was below3.50cm;the maximum deviation of course angle was3.87°;the mean absolute deviation of course angle was below1.7°;the maximum swing angle of the front wheels was3°in the steady state,and the corresponding standard deviation was0.80°.Therefore,the speed-adaption-based tractor navigation system proposed in this study was experimentally verified and could achieve the basic speed-adaptive tracking control and further improve the precision of tractor navigation.Furthermore,it is possible to meet the agronomic requirements of actual operations in the field.The approach proposed in this study provides potential theoretical guidance and certain significance for practical tractor field operations.