基于北斗卫星导航的拖拉机辅助驾驶系统研究

发布时间：2018-04-04 03:26

本文选题：北斗卫星导航　切入点：自动转向　出处：《江苏大学》2017年硕士论文

【摘要】：拖拉机辅助驾驶技术是农业机械自动化与智能化的发展要求,对实现精准农业具有重要意义。拖拉机辅助驾驶系统不仅可以减轻驾驶员的工作强度,还可以提高作业精度和生产效率。本文对拖拉机辅助驾驶技术中自动转向和路径跟踪算法进行了研究,以DF904拖拉机为研究对象,结合北斗卫星高精度定位设备,采用PID算法控制实现自动转向、变论域模糊算法控制实现路径跟踪。且通过遗传算法对PID参数值进行了优化,通过仿真和试验进行了验证。主要研究工作和结果具体如下:(1)确定了拖拉机自动导航系统的总体方案以及硬件选型,卫星导航辅助驾驶系统由路径跟踪系统和自动转向系统两部分组成。根据这两个系统的功能要求,分别进行了硬件的选型及集成。(2)自动转向控制算法设计。通过分析台架试验数据,获得拖拉机转向数学模型,根据转向模型,在Matlab环境下通过遗传算法确定了合适的PID参数值,并在Matlab/Simulink环境下进行了转向仿真试验,从理论上验证了采用遗传算法优化PID参数以控制自动转向的精确性。(3)路径跟踪控制算法设计。针对在农田坑洼不平、路面复杂实际状况下,拖拉机按预定路径行走时具有较强的非线性和不确定性,以及在路径偏差较小时会出现控制精度较低等情况,本文在模糊控制的基础上引入变论域模糊控制理论,利用模糊控制器来动态改变输入输出论域,即在模糊规则不变的情况下让初始论域随输入的变化而变化,解决了模糊控制在“零点”(偏差很小)附近控制精度低的问题。并且在Matlab/Simulink环境下进行了路径跟踪仿真试验,仿真结果表明,变论域模糊控制能有效地对拖拉机行驶进行路径跟踪。(4)试验研究。试验包括台架试验和路径跟踪试验。其中台架试验分别是在未加任何算法控制下以及采用PID算法控制下进行试验,通过试验确定了拖拉机的转向模型和PID控制下的转向性能。路径跟踪试验分别为在平坦路面和坑洼路面上,行驶速度为3km/h、4km/h时的直线路径跟踪,试验表明,在平坦路面上横向位置平均误差均小于3cm;在坑洼路面上遇障碍偏航后拖拉机能快速、准确地纠正偏差,返回预定路径。本文完成基于北斗导航的拖拉机辅助驾驶系统的相关研制,并通过仿真和试验测试了系统的性能。试验表明,该卫星导航辅助驾驶系统可实现拖拉机按预定路径自动行驶。
[Abstract]:Tractor-assisted driving technology is the development requirement of agricultural machinery automation and intelligence, and it is of great significance to achieve precision agriculture.Tractor-assisted driving system can not only reduce the driver's working intensity, but also improve the working precision and production efficiency.In this paper, the automatic steering and path tracking algorithms in tractor assisted driving technology are studied. Taking DF904 tractor as research object and combining Beidou satellite high precision positioning equipment, PID algorithm control is used to realize automatic steering.Variable domain fuzzy algorithm control to achieve path tracking.The parameters of PID are optimized by genetic algorithm and verified by simulation and experiment.The main research work and results are as follows: 1) the overall scheme and hardware selection of tractor automatic navigation system are determined. The satellite navigation assistant driving system is composed of two parts: path tracking system and automatic steering system.According to the functional requirements of the two systems, the hardware selection and integration. 2) automatic steering control algorithm are designed.The mathematical model of tractor steering is obtained by analyzing the data of bench test. According to the steering model, the appropriate PID parameter value is determined by genetic algorithm in Matlab environment, and the steering simulation test is carried out under Matlab/Simulink environment.The design of path tracking control algorithm using genetic algorithm to optimize PID parameters to control the accuracy of automatic steering is theoretically verified.In view of the fact that the tractor has strong nonlinearity and uncertainty when walking according to the predetermined path under the condition of uneven farmland potholes and complicated road surface, and the control precision will be low when the path deviation is small, etc.In this paper, the variable domain fuzzy control theory is introduced on the basis of fuzzy control, and the fuzzy controller is used to dynamically change the input and output domain, that is to say, the initial domain changes with the change of input when the fuzzy rules are invariant.The problem of low precision of fuzzy control near "zero" (small deviation) is solved.The simulation results of path tracking in Matlab/Simulink environment show that variable theory field fuzzy control can effectively track the path of tractor.The test includes bench test and path tracking test.The bench test is carried out under the control of no algorithm and the control of PID algorithm respectively. The steering model of tractor and the steering performance under PID control are determined by test.The path tracking tests are linear path tracking at a speed of 3 km / h and 4 km / h on flat and pothole pavement, respectively. The test results show that,The average lateral position error on the flat pavement is less than 3 cm, and the tractor can correct the deviation quickly and accurately and return to the predetermined path after yawing on the pothole pavement.In this paper, the development of tractor assisted driving system based on Beidou navigation is completed, and the performance of the system is tested by simulation and test.The test results show that the satellite navigation aided driving system can automatically drive the tractor according to the predetermined path.
【学位授予单位】：江苏大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：S219

【参考文献】