多智能体系统在分布式采样控制下的动力学行为  

Dynamic Behavior of Multi-agent Systems with Distributed Sampled Control

作　　者：余宏旺[1] 郑毓蕃[2] 

机构地区：[1]南京审计学院数学与统计学院,南京211815 [2]上海大学系统科学所,上海200436

出　　处：《自动化学报》2012年第3期357-365,共9页Acta Automatica Sinica

基　　金：国家自然科学基金(60674046);南京审计学院微分方程理论与应用(202720035)资助~~

摘　　要：本文运用图论、矩阵分析和现代控制理论等工具,研究信息网络传输下多智能体系统的协调控制和动力学行为.假设每个智能体通过数字化网络传感器获得其邻近智能体的位置状态,而且多智能体系统采取分布式线性控制协议.每个智能体被描述为一个简单的采样系统,多智能体系统转化为混合动力学系统.研究结果表明,多智能体系统所呈现的渐近聚集、周期振荡和发散动力学行为不仅和网络结构的代数特征有关,而且和每个智能体的动力学方程、采样周期有关.本文给出了具体精确的代数判据.仿真例子进一步验证了本文结果的有效性.This paper devotes to the study of coordinated control and dynamical behavior of multi-agent systems in digital communication network by means of graph theory, matrix analysis, and control theory. It is assumed that each agent of the system is described by a sampling dynamical system and is able to obtain the information of location coordinates of its neighbors by remote sensors. The system behavior depends on a linear distributed protocol in digital communication network. Thus, the whole multi-agent system can be considered as a hybrid dynamic system. Our results show that the multi-agent systems under consideration may appear three different dynamical behaviors, i.e., asymptotical aggregation, periodic oscillation, and divergence of agents. The dynamical behavior of the system depends not only on the algebraic characterization of the network topology but also on the dynamics of agents, as well as the sampling time period. Moreover, we give some algebraic criteria relevant to the system behavior. Examples show the effectiveness of our results.

关 键 词：多智能体 分布式控制协议 采样系统 一致性 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]