汽车工程

基于线性扩张状态观测器的平行泊车路径跟踪

展开
  • 1.南京航空航天大学 能源与动力学院,江苏 南京 210016; 2.清华大学 汽车安全与节能国家重点实验室,北京 100084
王健(1986-),男,博士生,主要从事智能驾驶、主动安全研究.E-mail:wangjian1987228@163.com

收稿日期: 2014-03-21

  修回日期: 2014-09-17

  网络出版日期: 2014-11-17

基金资助

国家自然科学基金资助项目(11072106, 51005133, 51375009)

Path Tracking for Parallel Parking Based on Linear Extended State Observer

Expand
  • 1.College of Energy and Power Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,Jiangsu,China; 2.State Key Laboratory of Automotive Safety and Energy,Tsinghua University,Beijing 100084,China
王健(1986-),男,博士生,主要从事智能驾驶、主动安全研究.E-mail:wangjian1987228@163.com

Received date: 2014-03-21

  Revised date: 2014-09-17

  Online published: 2014-11-17

Supported by

国家自然科学基金资助项目(11072106, 51005133, 51375009)

摘要

为了消除外界干扰和转向系统运动学模型的不确定性的影响,建立了平行泊车 系统的车辆运动学模型,设计了一个三阶线性扩张状态观测器,该观测器可将外界干扰和模型不确定性看作系统总的扰动量进行观测和补偿,而不需要建立被控对象的精确数学模型.基于该观测器,设计了平行泊车路径跟踪控制器,并对其性能进行了仿真和实车验证.仿真结果表明,所设计的平行泊车路径跟踪控制器的控制效果优于传统 PID 控制器,抗外界干扰能力更强.实车试验结果表明,该路径跟踪控制器能够精确控制车辆完成平行泊车任务,最大误差仅为0.111m.

本文引用格式

王健 赵又群 季学武 刘亚辉 臧利国 . 基于线性扩张状态观测器的平行泊车路径跟踪[J]. 华南理工大学学报(自然科学版), 2014 , 42(11) : 70 -77 . DOI: 10.3969/j.issn.1000-565X.2014.11.011

Abstract

In order to eliminate the external disturbance and remove the influence of the uncertainty of the steering system kinematics model,a vehicle kinematics model of parallel parking system is constructed and a third- order li- near extended state observer is designed.The observer takes the external disturbance and the model uncertainty for a total disturbance to observe and compensate without building the accurate mathematical model of controlled ob- jects.Based on this observer,a path- tracking controller is established for parallel parking and its control perfor- mance is verified by a simulation and a real vehicle test.Simulation results show that the proposed path- tracking controller shows a better performance and a stronger ability to resist the external disturbance in comparison with those of the traditional PID controller.The real vehicle test results show that the proposed path- tracking controller can control vehicles to finish parallel parking tasks precisely with a maximum error of only 0.111m.

文章导航

/