复杂动态网络控制研究进展  

RECENT PROGRESS IN CONTROL OF COMPLEX DYNAMICAL NETWORKS

作　　者：汪小帆[1] 苏厚胜[1] 

机构地区：[1]上海交通大学自动化系,上海200240

出　　处：《力学进展》2008年第6期751-765,共15页Advances in Mechanics

基　　金：国家自然科学基金(60674045;60731160629;70431002);上海市优秀学科带头人计划(07XD14017)资助项目~~

摘　　要：研究复杂网络结构性质与模型的主要目的之一就是为了了解网络结构与网络功能之间的关系,并在此基础上考虑改善网络性能的有效途径.综述了近年关于利用分布式控制的方法使得一个动态网络具有期望行为的一些研究进展.对于具有固定和连通的拓扑结构的复杂动态网络,牵制控制策略的有效性与网络拓扑密切相关.综述了牵制控制的可行性、稳定性分析以及控制策略比较研究.另一方面,对于具有时变拓扑结构的动态网络的控制,着重综述了移动多自主体网络系统的蜂拥控制,并特别阐述了如何把牵制控制的思想用于蜂拥控制.One of the main purposes in the study of complex network topology and modeling is to understand the influence of network structure on its function and consequently to find out an effective way to improve the network performance. This paper surveys recent advances in the distributed control approaches for making a dynamical network to behavior in a desired way. For a large-scale dynamical network with fixed and connected topology, great efforts have been focused on pinning control, i.e. to achieve the goal of control by directly adding control input to a fraction of nodes selected from the network, the feasibility, stability and strategy of which are reviewed in this paper. On the other hand, for a dynamical network with time-varying and possibly disconnected topology, flocking control of mobile multi-agent network systems with limited communication capabilities is reviewed, with emphasis laid on how the idea of pinning control can be applied to the flocking control.

关 键 词：复杂网络 控制 稳定性 牵制控制 蜂拥控制 

分 类 号：TP393[自动化与计算机技术—计算机应用技术] TJ768.28[自动化与计算机技术—计算机科学与技术]