计算机工程与应用 ›› 2020, Vol. 56 ›› Issue (18): 63-68.DOI: 10.3778/j.issn.1002-8331.1908-0281

• 理论与研发 • 上一篇    下一篇

碰撞锥检测改进的多智能体避障算法

许元云,何明,刘锦涛,周波,杨铖   

  1. 中国人民解放军陆军工程大学 指挥控制工程学院,南京 210007
  • 出版日期:2020-09-15 发布日期:2020-09-10

Multi-agent Obstacle Avoidance Algorithm Improved by Collision Cone Detection

XU Yuanyun, HE Ming, LIU Jintao, ZHOU Bo, YANG Cheng   

  1. College of Command and Control Engineering, The Army Engineering University of PLA, Nanjing 210007, China
  • Online:2020-09-15 Published:2020-09-10

摘要:

避障是多智能体能够适应复杂环境并顺利完成任务的必要条件之一。为使多智能体更快通过障碍物并达到一致,提出了一种多智能体避障控制算法。算法引入了避障系数,该系数由基于角度比较的碰撞锥检测方法来确定,并通过牵制控制输入完成多智能体的避障。证明了在该算法作用下所有智能体最终会避开障碍,避免碰撞并最终达到一致。通过实验仿真分析和对比,该方法能够使得多智能体更快避开障碍物。

关键词: 多智能体, 避障, 一致性, 碰撞锥检测

Abstract:

Obstacle avoidance is one of the necessary conditions for multi-agents to adapt to complex environments and successfully complete tasks. In order to make multi-agents pass obstacles faster and achieve consensus, a multi-agent obstacle avoidance control algorithm is proposed. An obstacle avoidance coefficient is proposed, which is determined by the collision cone detection method based on angle comparison, and the obstacle avoidance of the agent is completed by the control input and the obstacle avoidance coefficient. It is proved that all the agents in the algorithm eventually avoid obstacles, avoid collisions and reach consensus. Through experimental simulation analysis and comparison, this method can make multi-agents pass obstacles more quickly.

Key words: multi-agent, obstacle avoidance, consensus, collision cone detection