首页|切换拓扑下多智能体时变编队控制方法

切换拓扑下多智能体时变编队控制方法

Time-varying formation control method for multi-agent under switching topologies

扫码查看
针对通信存在测量噪声和链路频繁断开下的多智能体时变编队稳定控制问题,基于期望编队状态、智能体自身状态和通信拓扑,提出一种切换拓扑下多智能体时变编队控制方法.首先,通过矩阵替换对系统进行降维处理.然后,应用随机微分方程理论和 Lyapunov 稳定性定理推导出编队系统在普通切换和随机切换下稳定的充要条件以及编队收敛指标.最后,在四个智能体构成的编队和四个备用拓扑的仿真条件下,验证了所提方法可有效保证多智能编队在切换拓扑下的稳定性.相较于普通切换拓扑,随机切换拓扑能更快地将编队误差收敛至 10-2 以内,效率提高约 34.4%.同时编队收敛指标也证实了这一结论,且编队的精确性以及应用范围也要优于普通切换拓扑.
A time-varying formation control method for multi-agent under switching topologies is proposed,which utilizes the desired formation state,the agent itself state and the communication topology to achieve time-varying formation stability control for multi-agent with observation noise and frequent disconnection of communication links.Firstly,the dimensionality of the system is reduced by matrix replacement.Then,the stochastic differential equation theory and Lyapunov stability theorem are applied to derive necessary and sufficient conditions to ensure the stability of the formation system under normal switching and random switching,as well as the formation convergence index that can reflect the convergence performance of the formation.Finally,under the simulation conditions of a formation composed of four agents and four alternate topologies,it is verified that the proposed control method can effectively ensure the stability of the multi-agent formation in the switching topology.Compared with the normal switching topology,the random switching topology can converge the formation error to less than 10-2 faster,and the efficiency is increased by about 34.4%.At the same time,the convergence index of formation also confirms this conclusion and the accuracy of formation and application range is better than that of normal switching topology.

agentformation controlswitching topologystochastic differential equationLyapunov theorem

庞强伟、朱拥勇、陈晔、李宗吉、董理、黄成果

展开 >

中国人民解放军 92116 部队,葫芦岛 125003

海军工程大学 兵器工程学院,武汉 430000

智能体 编队控制 切换拓扑 随机微分方程 Lyapunov定理

湖北省重点研发计划海军工程大学创新基金

2021BID006HGCXJJ202109

2024

中国惯性技术学报
中国惯性技术学会

中国惯性技术学报

CSTPCD北大核心
影响因子:0.792
ISSN:1005-6734
年,卷(期):2024.32(5)