具有变通信时延的多智能体系统的平均一致性  

Average consensus of multi-agent systems with variable communication delays

作　　者：孙凤兰[1] 关治洪[1] 

机构地区：[1]华中科技大学控制科学与工程系,湖北武汉430074

出　　处：《华中科技大学学报（自然科学版）》2011年第7期71-75,共5页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：国家自然科学基金资助项目(61073025;60973012;60973039;60873021)

摘　　要：把一致性增益引入到一致性协议中,利用线性矩阵不等式(LMI)方法,给出了具有时变通信延迟的多智能体系统达到一致的充分条件.分析了当多智能体系统的网路拓扑是强连通平衡图且通信时延具有合适的上界时,带有有界时变延迟的多智体系统的平均一致性问题.证明了当在协议中引入一致性增益时,对于具有不同通信时延的多智能体系统,若时变通信时延的上界满足一定条件,则不论网络拓扑是不变的还是切换的,系统都能够达到平均一致,并且可通过求解线性矩阵不等式得到时变时延的上界.对于具有变时延的多智能体系统,存在相应的时延上界使得系统达到一致.所得结论对于具有时变拓扑网络的多智能体系统也成立.On the basis of consensus-gain and linear matrix inequality （LMI） technique, some suffi- cient conditions were obtained for the proposed consensus protocol of the multi-agent system with time-varying communication delays. It was proved that the multi-agent system with variable communi- cation delays can be achieved to average consensus under the fixed and switching network, if the net- work topology of the system is a strongly connected and balanced graph, and the communication delay has an appropriate upper bound. Furthermore, the greatest upper bound of time-varying communica- tion delays can be obtained by solving a simple feasible linear matrix inequality. The theoretical results can be extended to the systems with variable topology network.

关 键 词：网络 多智能体系统 时变时延 一致性增益 平均一致性 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]