Formation Control of Multi-agent System with Local Interaction and Artificial Potential Field
Luoyin Zhao, Zheping Yan, Yuqing Wang, Raye Chen-Hua Yeow
TL;DR
This work addresses formation control for multi-agent systems operating in cluttered environments. It introduces a local interaction control method (LICM) that fuses leader-follower and consensus frames (LFCF) and embeds a stress response mechanism with an artificial potential field (SRM-APF) for robust obstacle and collision avoidance. The paper proves asymptotic stability under infinitesimal rigidity via a Lyapunov analysis and demonstrates the method through triangle, square, and hexagon formations, showing effective navigation and formation maintenance. The approach reduces communication bottlenecks, enhances formation flexibility, and provides a practical framework for safe cooperative operation of MAS in uncertain environments.
Abstract
A novel local interaction control method (LICM) is proposed in this paper to realize the formation control of multi-agent system (MAS). A local interaction leader follower (LILF) structure is provided by coupling the advantages of information consensus and leader follower frame, the agents can obtain the state information of the leader by interacting with their neighbours, which will reduce the communication overhead of the system and the dependence on a single node of the topology. In addition, the artificial potential field (APF) method is introduced to achieve obstacle avoidance and collision avoidance between agents. Inspired by the stress response of animals, a stress response mechanism-artificial potential field (SRM-APF) is proposed, which will be triggered when the local minimum problem of APF occurs. Ultimately, the simulation experiments of three formation shapes, including triangular formation, square formation and hexagonal formation, validate the effectiveness of the proposed method.