计算机应用 ›› 2010, Vol. 30 ›› Issue (8): 2017-2020.

• 人工智能 • 上一篇    下一篇

极坐标下基于迭代学习的移动机器人轨迹跟踪控制

阎世梁1,张华1,王银玲2,肖晓萍3   

  1. 1. 西南科技大学
    2.
    3. 西南科技大学工程技术中心
  • 收稿日期:2010-02-04 修回日期:2010-03-03 发布日期:2010-07-30 出版日期:2010-08-01
  • 通讯作者: 阎世梁
  • 基金资助:
    国家自然科学基金资助项目;西南科技大学国防重点学科实验室重点培育项目;西南科技大学青年基金

Trajectory tracking of mobile robot using iterative learning control in polar coordinates

  • Received:2010-02-04 Revised:2010-03-03 Online:2010-07-30 Published:2010-08-01

摘要: 为提高自主移动机器人对一类特殊轨迹的重复跟踪能力,在极坐标下建立了3轮全向移动机器人的运动学模型,结合离散时域下对轨迹跟踪问题的描述方法,采用开闭环P型迭代学习控制算法,并在给定条件下证明了其收敛性,随着迭代次数的增加,该算法能够有效改善动态不确定环境中系统的稳定性与收敛的快速性。通过将仿真结果作用于实际动态系统的初始控制输入,从而在实际环境下能以较少的迭代过程来获取控制律。实验结果表明,在仿真环境下机器人可以较好地跟踪玫瑰曲线,在实际机器人测试中,机器人能够较好地跟踪期望轨迹,从而证实了该方法对提高自主移动机器人轨迹跟踪能力的可行性与有效性。

关键词: 移动机器人, 运动学, 轨迹跟踪, 学习算法

Abstract: To improve the repetitive tracking performance of autonomous mobile robot for a class of special curves, the robots kinematics model was studied in polar coordinates; an openclosedloop Ptype iterative learning control (ILC) algorithm which is described in discretetime domain was applied, and the convergence proof of the ILC algorithm was presented under the given conditions. With the increasing iteration times, a better stabilization and convergence rate of the system can be achieved under dynamic uncertain environment. By taking the simulation control law as the initial control inputs on real robot platform firstly, the real control law can be obtained with less iterations under practical environment. The simulation results show good performance for tracking and the real experimental results verify the feasibility and efficiency on actual mobile robot platform.

Key words: mobile robot, kinematics, trajectory tracking, learning algorithm