通信延时和数据丢包下事件驱动的多智能体系统一致性研究（英文）  

Event-triggered consensus of multi-agent systems with data transmission delays and random packet dropouts

作　　者：崔彦良 费敏锐[1] 杜大军[1] 李慷[2] 

机构地区：[1]上海市电站自动化技术重点实验室,上海大学机电与自动化学院,上海200072 [2]英国贝尔法斯特女王大学电子电气与计算机学院,英国贝尔法斯特BT95AH

出　　处：《控制理论与应用》2015年第9期1208-1218,共11页Control Theory & Applications

基　　金：Supported by Key Project of Science and Technology Commission of Shanghai Municipality(14JC1402200,15JC1401900);National Key Scientific Instrument and Equipment Development Project(2012YQ15008703);National Science Foundation of China(61473182);Shanghai Rising-Star Program(13QA 1401600)

摘　　要：本文研究受网络通信延时和数据随机丢包的多智能系统一致性问题,探索事件驱动的分布式协同控制策略.首先针对两类普遍应用的事件触发器,提出了一个可用于选择触发策略的触发频率比较方法.然后提出了分布式协同控制律以保证系统的渐近一致性,并给出了相应的时滞依赖Markov切换控制器设计新方法.本文所提的控制策略不仅保证系统一致性目标,而且能显著减少通信数据传输量并降低控制器计算负担.最后,通过仿真算例验证了所提方法的有效性.This paper investigates the event-triggered consensus of multi-agent systems（MASs）with time delay and random packet dropout in data transmission.The event-triggered scheme is employed for broadcasting fewer necessary data to the neighbor agents through communication networks only when its threshold is violated.Based on prior topology information of the MASs,an approximate frequency comparison method is firstly proposed to choose the suitable one from two typical event-triggers.For guaranteeing the asymptotical consensus of MASs as well as enhancing system robustness against the communication drawbacks,a distributed Markov switching controller is designed.The sufficient delay dependent stability conditions are obtained and the corresponding controller design methods are subsequently presented.With the proposed strategy,it is shown that the amount of communication packages and the controller updates can be significantly reduced without introducing any significant negative effect on the consensus.Finally,the effectiveness of the proposed theoretical approach is validated through several numerical examples.

关 键 词：事件驱动 渐近一致性 多智能体系统 通信延时 数据丢包 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]