线性多智能体系统在固定和切换拓扑下的领航跟随控制  

Leader-following consensus of linear multi-agent systems under fixed and switching topologies

作　　者：宋莉[1] 伍清河[1] 王垚[1] 王寅秋[1] 

机构地区：[1]北京理工大学自动化学院,北京100081

出　　处：《控制与决策》2013年第11期1685-1690,1696,共7页Control and Decision

基　　金：国家自然科学基金项目(61074031)

摘　　要：研究线性多智能体系统的领航跟随一致性问题.假设每个多智能体系统只能得到其邻域的输出测量信息,在此条件下,讨论多智能体在有向固定网络拓扑和无向切换网络拓扑两种情况下的一致性问题.针对这两种情况,提出含有一种分布式观测器的一致性控制算法.应用Lyapunov稳定性理论证明了若单个多智能体系统是可镇定和可检测的,且网络连接拓扑只需满足简单的结构,则系统能够达到领航跟随一致性.仿真结果验证了理论分析的正确性.The leader-following consensus problem of multi-agent systems with general linear systems is studied. It is assumed that each agent has only access to the relative output measurements with respect to its neighbors. Two cases are considered： the directed network with fixed topology and the undirected networks with switching topologies. For both of cases, a distributed observer-type consensus protocol only based on output measurements is proposed. It is proved that the leader-following consensus can be achieved only assuming stabilizability and detectability of the individual systems and imposing very mild connectedness assumption on the interconnection structure by using Lyapunov theory. Finally, simulation results show the effectiveness of the theoretical results.

关 键 词：多智能体系统 领航跟随 观测器型 代数RICCATI方程 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]