首页 >

切换多智能体系统的全分布式比例一致性控制

孙亚平1,王勇2,袁爽1,苏厚胜2,妙锁霞3,杨鑫松1   

  1. 1. 四川大学
    2. 华中科技大学
    3. 江西水利电力大学
  • 收稿日期:2025-10-30 修回日期:2026-01-04 出版日期:2026-01-09 发布日期:2026-01-09
  • 通讯作者: 杨鑫松
  • 基金资助:
    国家重点研发计划;国家自然科学基金;特殊环境机器人技术四川省重点实验室开放基金

Fully distributed scaled consensus for switched multi-agent systems

  • Received:2025-10-30 Revised:2026-01-04 Online:2026-01-09 Published:2026-01-09
  • Contact: Xinsong Yang

摘要: 本文对切换多智能体系统的全分布式比例一致性控制问题进行了研究,其中多智能体系统同时考虑了个体模型和通信网络的切换特性;控制输入受饱和约束限制;系统状态受比例特性影响;所采用的随机切换规则兼具马尔可夫切换与驻留时间切换的双重优势。由于个体模型与通信网络的随机切换特性、通信网络的全局信息未知、输入饱和约束以及个体状态间的比例特性所带来的难点,全分布式的比例一致性问题很难直接用现有方法解决。因此,本文设计了新的自适应比例一致性算法和李亚普诺夫函数,引入了代数黎卡提方程,并基于此给出了切换多智能体系统能实现几乎必然的全分布式指数比例一致性的条件。特别地,所获条件具有可行解且极易求解。最后,引入无人机集群验证了所提控制方案的有效性。

关键词: 多智能体系统, 一致性, 随机切换, 全分布式, 自适应控制, 输入饱和

Abstract: This paper investigates the fully distributed scaled consensus problem of switched multi-agent systems, in which the MAS considers both agent kinematics and communication network switching; control inputs are limited by saturation constraint; proportion property has an effect on system states; and a stochastic switching rule that combines the advantages of Markov switching and dwell-time switching is adopted. The challenges include random switching, lack of global communication information, input saturation, and scaled state constraints, making existing methods inapplicable. Therefore, a new adaptive consensus algorithm and Lyapunov function are proposed, and the algebraic Riccati equation is introduced. Based on this, almost surely exponential consensus conditions that have a feasible solution and are easy to obtain are obtained. Finally, the proposed control scheme is validated using an unmanned aerial vehicle swarm.

Key words: multi-agent systems, consensus, random switching, fully distributed, adaptive control, input saturation

中图分类号: