中国机械工程 ›› 2010, Vol. 21 ›› Issue (04): 491-495.

• 车辆工程 • 上一篇    下一篇

用于汽车驾驶机器人的车辆性能自学习方法

陈刚;张为公;龚宗洋;孙伟
  

  1. 东南大学,南京,210096
  • 出版日期:2010-02-25 发布日期:2010-03-10
  • 基金资助:
    高等学校博士学科点专项科研基金资助项目(200802861061);江苏省汽车工程重点实验室开放基金资助项目(QC200603)
    Specialized Research Fund for the Doctoral Program of Higher Education of China(No. 200802861061)

A Vehicle Performance Self Learning Method Applied to Robot Driver

Chen Gang;Zhang Weigong;Gong Zongyang;Sun Wei
  

  1. Southeast University,Nanjing,210096
  • Online:2010-02-25 Published:2010-03-10
  • Supported by:
    Specialized Research Fund for the Doctoral Program of Higher Education of China(No. 200802861061)

摘要:

为了缩短在进行汽车试验前驾驶机器人对不同车型的适应性调整时间,提出了一种用于驾驶机器人的车辆性能自学习方法,对影响驾驶机器人驾驶行为的车辆尺寸和汽车性能参数进行自学习。车辆尺寸的学习通过示教再现实现,汽车性能自学习中油门和制动执行器的指令信号通过所需的车辆驱动功率来确定。对因长时间驾驶引起的控制参数变化进行在线优化,以补偿长时间试验过程中汽车零部件的磨损。试验结果表明,提出的方法实现了驾驶机器人的自学习、自适应、自补偿,驾驶机器人具有良好的车型适应能力,车速跟踪精度满足试验的要求,能消除汽车试验中人为因素的影响。

关键词:

Abstract:

To reduce the preparation time before vehicle test,a vehicle performance
self learning method applied to robot driver was proposed.The vehicle geometric parameters and vehicle performance parameters which influence the driving behavior of robot driver were self learned. The vehicle geometric parameters were learned by key box, and the command signals applied to the accelerator and brake actuators were determined by the power needed to drive the vehicle during the process of vehicle performance parameter self learning. Furthermore, the capability of control parameter self-optimization on line made it possible to compensate for the wear of vehicle during the long term test. Experimental results demonstrate that the self-learning, self-adaptation and self-compensation are realized by the proposed method and the robot driver can be applicable to a wide variety of vehicles. Moreover, the accuracy of vehicle speed tracking meets the requirements of vehicle test. Therefore, the method provides a solution to eliminate the uncertainty of vehicle test results by human drivers.

Key words: vehicle, robot driver, vehicle performance parameter, self learning, self compensation

中图分类号: