基于距离及相对角度信息的多智能体系统刚性编队任务规划

Translated title of the contribution: Rigid Formation Task Planning for Multi-Agent Systems Based on Distance and Relative Angle Information

Huang Jie*, Li Junjie

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

This paper proposes two rigid formation task planning schemes for multi-agent systems to address the problem of maintaining rigid formation task planning under the condition of only obtaining local distance and relative angle information of multi-agent systems. Firstly, a rigid formation planning scheme based on distance and relative angle is proposed to ensure the rigid formation of intelligent agents in unknown relative positions. Secondly, the task planning method of null space projection is used to solve the conflict between formation tasks and obstacle avoidance tasks, ensuring the stability of formation during the movement process. An overall rigid formation planning scheme is proposed that combines distance and relative angle again, and the multi-agent system is considered as a whole for obstacle avoidance during the obstacle avoidance process, ensuring stable rigid formation and the completion of obstacle avoidance tasks. After simulation verification, the two methods proposed in this paper can maintain a rigid formation and complete obstacle avoidance tasks well under the condition of weakening relative position information. Compared with the null space behavior control method that relies on relative position, it weakens the dependence on relative position information.

Translated title of the contributionRigid Formation Task Planning for Multi-Agent Systems Based on Distance and Relative Angle Information
Original languageChinese (Traditional)
Pages (from-to)79-90
Number of pages12
JournalAero Weaponry
Volume31
Issue number2
DOIs
Publication statusPublished - 30 Apr 2024
Externally publishedYes

Fingerprint

Dive into the research topics of 'Rigid Formation Task Planning for Multi-Agent Systems Based on Distance and Relative Angle Information'. Together they form a unique fingerprint.

Cite this