中国物理B ›› 2009, Vol. 18 ›› Issue (10): 4222-4228.doi: 10.1088/1674-1056/18/10/023

• • 上一篇    下一篇

Collision avoidance for a mobile robot based on radial basis function hybrid force control technique

温淑焕   

  1. Key Lab of Industrial Computer Control Engineering of Hebei Province, Yanshan University, Qinhuangdao 066004, China
  • 收稿日期:2008-12-07 修回日期:2009-04-16 出版日期:2009-10-20 发布日期:2009-10-20
  • 基金资助:
    Project supported by the Science and Technology Stress Projects of Hebei Province, China (Grant No 07213526).

Collision avoidance for a mobile robot based on radial basis function hybrid force control technique

Wen Shu-Huan(温淑焕)†   

  1. Key Lab of Industrial Computer Control Engineering of Hebei Province, Yanshan University, Qinhuangdao 066004, China
  • Received:2008-12-07 Revised:2009-04-16 Online:2009-10-20 Published:2009-10-20
  • Supported by:
    Project supported by the Science and Technology Stress Projects of Hebei Province, China (Grant No 07213526).

PDF (PC)

6173

赞

摘要： Collision avoidance is always difficult in the planning path for a mobile robot. In this paper, the virtual force field between a mobile robot and an obstacle is formed and regulated to maintain a desired distance by hybrid force control algorithm. Since uncertainties from robot dynamics and obstacle degrade the performance of a collision avoidance task, intelligent control is used to compensate for the uncertainties. A radial basis function (RBF) neural network is used to regulate the force field of an accurate distance between a robot and an obstacle in this paper and then simulation studies are conducted to confirm that the proposed algorithm is effective.

Abstract: Collision avoidance is always difficult in the planning path for a mobile robot. In this paper, the virtual force field between a mobile robot and an obstacle is formed and regulated to maintain a desired distance by hybrid force control algorithm. Since uncertainties from robot dynamics and obstacle degrade the performance of a collision avoidance task, intelligent control is used to compensate for the uncertainties. A radial basis function (RBF) neural network is used to regulate the force field of an accurate distance between a robot and an obstacle in this paper and then simulation studies are conducted to confirm that the proposed algorithm is effective.

Key words: mobile robot collision avoidance, hybrid force/position control, path planning, RBF neural network

中图分类号:  (Neural networks)

  • 84.35.+i
45.40.Bb (Rotational kinematics) 45.40.Ln (Robotics)