en
×

分享给微信好友或者朋友圈

使用微信“扫一扫”功能。
作者简介:

邓志良,男,博士,教授,博士生导师,研究方向为智能识别与控制.dzl8188@qq.com

中图分类号:TP273

文献标识码:A

DOI:10.13878/j.cnki.jnuist.2023.02.011

参考文献 1
朱旭,张逊逊,闫茂德,等.基于一致性的无人机编队控制策略[J].计算机仿真,2016,33(8):30-34;ZHU Xu,ZHANG Xunxun,YAN Maode,et al.UAV formation control strategy based on consensus[J].Computer Simulation,2016,33(8):30-34
参考文献 2
刘建刚,杨胜杰.具有容性负载的直流微电网系统分布式协同控制[J].自动化学报,2020,46(6):1283-1290;LIU Jiangang,YANG Shengjie.Distributed cooperative control of DC micro-grid systems with capacitive loads[J].Acta Automatica Sinica,2020,46(6):1283-1290
参考文献 3
刘伟,李大卫,戴洪德,等.具有避障的机器人集群系统分布式编队控制[J].机械与电子,2021,39(7):44-48;LIU Wei,LI Dawei,DAI Hongde,et al.Distributed multi-robots system formation control with obstacle avoidance[J].Machinery & Electronics,2021,39(7):44-48
参考文献 4
涂志亮,王强,沈毅.一种目标监测的移动传感器网络覆盖分布式优化算法[J].控制与决策,2012,27(9):1353-1358;TU Zhiliang,WANG Qiang,SHEN Yi.A distributed coverage optimization algorithm for target monitoring in mobile sensor network[J].Control and Decision,2012,27(9):1353-1358
参考文献 5
Kia S S,Cortés J,Martínez S.Distributed convex optimization via continuous-time coordination algorithms with discrete-time communication[J].Automatica,2015,55:254-264
参考文献 6
Xi C G,Khan U A.Distributed subgradient projection algorithm over directed graphs[J].IEEE Transactions on Automatic Control,2017,62(8):3986-3992
参考文献 7
Wen X N,Qin S T.A projection-based continuous-time algorithm for distributed optimization over multi-agent systems[J].Complex & Intelligent Systems,2021:1-11
参考文献 8
Yi P,Hong Y G,Liu F.Distributed gradient algorithm for constrained optimization with application to load sharing in power systems[J].Systems & Control Letters,2015,83:45-52
参考文献 9
吴文,刘斌,姚靖,等.基于事件触发网络控制系统的分布式控制[J].湖南工业大学学报,2014,28(4):61-66;WU Wen,LIU Bin,YAO Jing,et al.Event-triggered based distributed control in networked control systems[J].Journal of Hunan University of Technology,2014,28(4):61-66
参考文献 10
Tabuada P.Event-triggered real-time scheduling of stabilizing control tasks[J].IEEE Transactions on Automatic Control,2007,52(9):1680-1685
参考文献 11
Dimarogonas D V,Frazzoli E,Johansson K H.Distributed event-triggered control for multi-agent systems[J].IEEE Transactions on Automatic Control,2012,57(5):1291-1297
参考文献 12
Liu Z X,Chen Z Q.Event-triggered average-consensus for multi-agent systems[C]//Proceedings of the 29th Chinese Control Conference(CCC’10),2010:4506-4511
参考文献 13
Xie D S,Xu S Y,Zhang B Y,et al.Consensus for multi-agent systems with distributed adaptive control and an event-triggered communication strategy[J].IET Control Theory & Applications,2016,10(13):1547-1555
参考文献 14
Fan Y,Feng G,Wang Y,et al.Distributed event-triggered control of multi-agent systems with combinational measurements[J].Automatica,2013,49(2):671-675
参考文献 15
Li H Q,Liu S,Soh Y C,et al.Achieving linear convergence for distributed optimization with zeno-like-free event-triggered communication scheme[C]//2017 29th Chinese Control and Decision Conference(CCDC).May 28-30,2017,Chongqing,China.IEEE,2017:6224-6229
参考文献 16
Du W,Yi X L,George J,et al.Distributed optimization with dynamic event-triggered mechanisms[C]//2018 IEEE Conference on Decision and Control.December 17-19,2018,Miami,FL,USA.IEEE,2018:969-974
目录contents

    摘要

    针对多智能体系统优化问题,提出一种基于动态事件触发机制的分布式优化算法.基于李雅普诺夫函数方法设计一种新型的动态事件触发控制器,相较于传统静态事件触发控制方法,所提出算法可有效降低多智能体间通信负担以及控制器计算负担.此外,利用周期采样信息进行事件触发条件设计,可避免智能体连续检测事件触发条件,并可消除Zeno现象.通过数值仿真验证了算法的有效性.

    Abstract

    The dynamic event triggered mechanism is used to design a distributed optimization algorithm for multi-agent systems.Compared with traditional static triggered control,the dynamic event triggered controller based on Lyapunov function can effectively reduce the communication burden between agents as well as the calculation burden of controllers.In addition,the event triggering condition is designed using periodic sampling information,thus is not required to be checked repeatedly by agents.Moreover,Zeno behavior can be avoided.A numerical simulation is given to verify the effectiveness of the algorithm.

  • 0 引言

  • 多智能体系统作为控制行业的前沿科技,在无人机编队[1]、微电网控制[2]、机器人群集[3]、无线传感器网络[4]等方面具有广泛应用,因此,多智能体系统的分布式优化问题受到大量研究者的关注[5-6],其研究目的是为通过分布式控制方法实现多智能体系统总成本函数最小化.

  • 文献[7-8]针对等式约束以及不等式约束下的优化问题提出了连续时间分布式优化算法.为实现系统最优,各智能体之间需要进行连续的信息交互,但实际系统中,由于网络的带宽有限,所设计算法很难满足实际应用.基于此,学者将事件触发控制方法应用于分布式优化问题,当智能体之间的状态达到触发条件时,智能体之间进行通信,反之,则不进行通信[9].依据事件触发条件所设计的算法可有效避免执行过程中智能体连续通信以及控制器连续更新问题.文献[10]针对通信约束下的控制问题,设计了一种简单的事件触发控制器,并证明所提出的调度策略可以保证半全局渐近稳定性; 文献[11-12]在文献[10]的基础上将事件触发机制应用到一阶系统的优化问题中,解决了传统周期采样控制智能体间通信频繁的问题,但是需要对事件触发条件进行连续检测; 文献[13]基于事件触发提出一种自适应控制策略,系统的触发时刻只与智能体自身的状态和邻居最新触发时刻的状态有关,避免了对邻居状态的连续检测; 文献[14]基于事件触发设计出一种组合测量方式,使得智能体只在自身事件触发时刻进行控制输入更新.由于对系统状态的逼近过多,利用系统的先验信息来估计下一个事件触发时间的自触发控制往往会引起控制器的更新,Zeno现象成为一个必须要讨论的问题,例如文献[15]通过利用离散周期采样序列对智能体进行检测,有效地避免了一阶离散系统出现Zeno现象.上述文献事件触发条件均为静态事件触发条件.文献[16]针对优化问题,提出一种基于动态事件触发的分布式优化算法,通过引入内部动态变量,设计了动态控制器触发条件,减少了系统的通信负担.但是其需要连续检测所提出的事件触发条件,且Zeno现象难以处理.

  • 受到文献[14-16]的启发,针对多智能体系统二次凸优化问题,本文设计出一种基于周期采样信息的分布式动态事件触发优化算法.该算法采用周期采样信息进行事件触发条件设计,两次触发时间的最小间隔为采样周期,可有效避免事件触发条件的连续检测问题以及Zeno现象,更符合实际系统运行机制.相较于传统的静态事件触发条件,所设计的动态事件触发条件触发频率更低,可有效降低智能体间通信频率以及控制器更新频率.

  • 1 代数图论及问题描述

  • 1.1 代数图论

  • 智能体之间的网络拓扑可用G=(VEA)表示,其中V={v1v2,···,vn}表示节点集,E={V×V}表示智能体之间的边集,A=[aij]∈Rn×n表示邻接矩阵.若(vivj)∈E,则aij=1,表示j可以接收到来自i的信息,否则aij=0.若图为无向图,则表示邻接矩阵对称,即aij=aji,不考虑无自环的情况,因此aii=0.入度矩阵D定义为diag{d1d2,···,dn},其中di=j=1n aij; 拉普拉斯矩阵L=[lij]n×n,其对角元素为lii=ii aij,非对角元素为lij=-aij.矩阵LDA满足关系L=D-A.

  • 若无向图G中任意两个节点之间都是连接的,则称图G为无向连通图,对于无向连通图,0是矩阵L的一个简单特征值,L的最大特征值λnL)<2dmax,其中dmaxn-1 且dmax=max{dii=1,2,···,n}.

  • 1.2 问题描述与分析

  • 本文主要研究多智能体系统二次凸优化问题,定义如下:

  • 1)问题描述:

  • mini=1n fimi
    (1)
  • fimi=αimi2+βimi+γi
    (2)
  • 2)约束条件:

  • s.t. i=1n mi=M,
    (3)
  • 其中:fimi)为智能体i的私有成本函数; αiβiγi为智能体i的成本参数且均为正常数; mi为智能体i成本函数fimi)的自变量.多智能体系统中,智能体i仅可获得其自身信息和邻居信息.

  • 基于拉格朗日乘子法可得:

  • f1m1m1=f2m2m2==fnmnmn=η*
    (4)
  • η*为拉格朗日乘子且为成本函数的优化解,由式(3)、(4)可进一步得到:

  • 2α1m1+β1==2αnmn+βn=η*
    (5)
  • 由式(2)、(5)可得:

  • η*=M+i=1n βi2αii=1n 12αi
    (6)
  • 本文的目标是设计出一个基于事件触发的分布式优化算法,可以求出问题(1)、(2)的最优解mi*,且可以降低系统的通信负担和控制器的更新频率,对应的优化值mi*

  • mi*=η*-βi2αi
    (7)
  • 定义变量

  • ηi(t)2αimi(t)+βi,
    (8)
  • 若满足条件:

  • η1(t)=η2(t)==ηn(t)=η*
    (9)
  • m1t=m2t==mnt=mi*,基于事件触发的二次凸优化问题(1)、(2)可转化为关于ηit)的一致性问题.

  • 2 分布式算法设计

  • 针对无向连通拓扑下的二次凸优化问题(1)、(2)设计算法如下:

  • ηi(t)=-2αij=1n aijη^i(t)-η^j(t)
    (10)
  • ηi(0)=2αimi(0)+βi
    (11)
  • mi(t)=ηi(t)-βi2αi,
    (12)
  • 其中:αiβi为正常数; aij为智能体的邻接矩阵; η^it为当前事件触发的状态值; mi(0)为智能体成本函数的初始状态,满足i=1n mi0=M.采样周期为h,当ttlitl+1iη^itηitlilN+tlil=1表示智能体i在采样时刻的序列,每个智能体的事件触发时刻为一个单调递增的序列.当ttlitl+1i时将测量偏差变量定义为

  • ei(t)=η^i(t)-ηi(t),
    (13)
  • 其中η^it=ηitkittkitk+1iηit表示智能体it时刻的状态,η^it表示智能体i最近一次触发时的状态值.

  • 定义

  • y^i2(t)=jNi+ aijη^i(t)-η^j(t)2,
    (14)
  • ξ˙i(t)=-μiξi(kh)+δiσy^i2(kh)-diei2(kh).
    (15)
  • 式(10)的动态触发条件可以设计为

  • θidiei2(kh)-σy^i2(kh)ξi(kh),
    (16)
  • 其中ξi(0)>0且di=dini),θiμiδiσ均为正参数.

  • 式(10)可用矩阵表示为

  • η˙(t)=-Kη(t)
    (17)
  • 其中:

  • η (t) =η1 (t) , η2 (t) , , ηn (t) T, η^ (t) =η^1 (t) , η^2 (t) , , η^n (t) T, e (t) =e1 (t) , , en (t) T, K=diag2α1, 2α2, , 2αn.

  • 定理1 假设多智能体系统通信拓扑图为无向连通图,算法(10)在事件触发条件(16)下若参数满足以下条件:

  • 12-2hα-λn-2σ>0
    (18)
  • hμi+δiθi<1,
    (19)
  • μi>1-δiθi
    (20)
  • 则优化问题(1)、(2)可求得最优解mi*t=η*-βi2αi,其中α-=maxαii=1,2nλnL的最大特征值.

  • 证明 设计如下李雅普诺夫函数

  • V(t)=i=1n 14αiδi2+i=1n ξi(t),
    (21)
  • 其中变量:

  • δi(t)=ηi(t)-η*
    (22)
  • 由式(15)和式(16)可知,当满足:

  • ξ˙i(t)-μiξi(kh)-δiθiξi(kh),
    (23)
  • 对于t∈[khkh+h)可得:

  • ξi(kh)1-hμi+δiθikξi(0)
    (24)
  • 由式(19)可知0<1-hμi+δiθi<1,故ξikh)>0,由式(23)、式(24)可得:

  • ξi(t)ξi(kh)-(t-kh)μi+δiθiξi(kh)1-hμi+δiθik+1ξi(0)>0,
    (25)
  • 故可得Vt)≥0,当t∈[khkh+h),tik=khkN,每个智能体的事件触发时刻均是采样周期的整数倍,每个智能体与其邻居仅在事件触发时刻进行数据交换.

  • V˙(t)=i=1n 12αiδiδ˙i+i=1n ξ˙i(t)=i=1n 12αiηi(t)-η*η˙i(t)-η˙*+i=1n -μiξi(kh)+δiσy^i2(kh)-diei2(kh),
    (26)
  • 其中:

  • V1=i=1n 12αiηi(t)-η*η˙i(t)-η˙*=-i=1n ηi(t)j=1n aijη^i(t)-η^j(t)+i=1n η*j=1n aijη^i(t)-η^j(t),
    (27)
  • V2=i=1n -μiξi(kh)+δiσy^i2(kh)-diei2(kh).
    (28)
  • 由拉普拉斯矩阵的对称性可知:

  • i=1n j=1n aijη^i(t)-η^j(t)=0
    (29)
  • t∈[khkh+h)时,η˙t=η˙kh,进而可得:

  • η(t)=(t-kh)η˙(kh)+η(kh)
    (30)
  • 因此:

  • V1=-i=1n ηi(t)j=1n aijη^i(t)-η^j(t)=-ηT(t)Lη^(t)=-((t-kh)η˙(kh)+η(kh))TLη^(t)=-((t-kh)(-KLη^(t))+η(kh))TLη^(t)=(t-kh)η^T(t)LTK(Lη^(t))-ηT(kh)Lη^(t).
    (31)
  • 将式(13)代入可得:

  • V1=(t-kh)η^T(t)LTK(Lη^(t))-(η^(kh)-e(kh))TLη^(t).
    (32)
  • 由于t-khh,故可得:

  • V1hη^T(t)LTK(Lη^(t))-η^T(kh)Lη^(t)+eT(kh)Lη^(t).
    (33)
  • 因为Lη^tTKLη^t2α-Lη^tTLη^t,其中α-=maxαii=1,2n

  • 故可得:

  • V12hα-η^T (t) LTLη^ (t) -η^T (kh) Lη^ (t) +eT (kh) Lη^ (t) 2hα-λnη^T (t) Lη^ (t) -η^T (kh) Lη^ (t) +

  • eT(kh)Lη^(t)
    (34)
  • η^T(t)Lη^(t)=12i=1n jNi aijη^i-η^j2=12i=1n y^i2,
    (35)
  • eT(kh)Lη^(t)=i=1n ei(kh)jNi+ aijη^i(t)-η^j(t)14i=1n jNH aijη^i(t)-η^j(t)2+i=1n diei2(kh)i=1n diei2(kh)+14y^i2(t).
    (36)
  • 将式(34)—(36)代入式(26)可得:

  • V˙(t)=V1+V22hα-λnη^T(t)Lη^(t)-η^T(kh)Lη^(t)+eT(kh)Lη^(t)+i=1n -μiξi(kh)+δiσy^i2(kh)-diei2(kh)-1-2hα-λnη^T(t)Lη^(t)+eT(kh)Lη^(t)+i=1n -μiξi(kh)+δiσy^i2(kh)-diei2(kh)-1212-2hα-λn-2σi=1n y^i2(t)+i=1n σy^i2(t)-diei2(kh)-1+δi+i=1n -μiξi(kh).
    (37)
  • 由式(16)可得:

  • σy^i2(kh)-diei2(kh)-1θiξi(kh)
    (38)
  • 将式(38)代入式(37)可得:

  • V˙(t)-1212-2hα-λn-2σi=1n y^i2(t)-i=1n δi-1θiξi(kh)-i=1n μiξi(kh)-1212-2hα-λn-2σi=1n y^i2(t)-i=1n δi-1θi+μiξi(kh).
    (39)
  • 当条件(18)—(20)满足时,(t)≤0.

  • V˙t=0时有:

  • ξi(t)=0
    (40)
  • 因此得到:

  • i=1n y^i2(t)=0
    (41)
  • η^1(t)=η^2(t)=η^3(t)==η^n(t)=c1,
    (42)
  • c1为未知常数,其值由后文给出.根据事件触发条件进一步得到:

  • e1(t)=e2(t)=e3(t)==en(t)=0,
    (43)
  • 因此得到

  • η1(t)=η2(t)==ηn(t)=c1
    (44)
  • 由式(10)可知:

  • i=1n 0t 12αiηi(τ)dτ=0t i=1n j=1n aijη^j(t)-η^i(t)dτ
    (45)
  • 由拉普拉斯矩阵的对称性可知

  • i=1n j=1n aijη^j (t) -η^i (t) =0,

  • 进而得到:

  • i=1n 0t 12αiη˙i(τ)dτ=0
    (46)
  • i=1n 12αiηi(t)-i=1n 12αiηi(0)=0
    (47)
  • i=1n 12αiηi(t)=i=1n 12αi2αimi(0)+βi,
    (48)
  • limt ηi(t)=η*=i=1n mi(0)+i=1n βi2αii=1n 12αi=M+i=1n βi2αii=1n 12αi.
    (49)
  • 根据ηit2αimit+βi可得:

  • limt mi(t)=mi*(t)=η*-βi2αi,
    (50)
  • 证毕.

  • 注1 该算法采用周期采样信息进行事件触发条件设计,两次触发时间的最小间隔为采样周期,可有效避免事件触发条件的连续检测问题以及Zeno现象.

  • 3 仿真实验

  • 首先给出一个由4个智能体组成的通信拓扑图(图1),并通过Matlab仿真来验证算法的可行性.

  • 每个智能体的成本函数用二次凸函数fizi=αimi2+βimi+γi来表示,其中各参数的设置如表1所示.

  • 图1 通信拓扑图

  • Fig.1 Communication topology

  • 表1 函数fimi)的成本函数

  • Table1 Cost function of function fi (mi)

  • 令初始值m1(0)=140,m2(0)=110,m3(0)=100,m4(0)=90,M=i=14 mi=440,则对应的η1(0)=27.08,η2(0)=16.04,η3(0)=21.20,η4(0)=14.96.由式(6)和式(7)可知对应的优化解η*1=η*2=η*3=η*4=η*=19.320,m*1=99.583,m*2=132.778,m*3=91.048,m*4=116.585.ξ(0)=[5,5,5,5],h=0.02 s,σ=0.1,θ1=θ2=θ3=θ4=1,δ1=δ2=δ3=δ4=0.2,K=[0.192,0,0,0; 0,0.144,0,0; 0,0,0.210,0; 0,0,0,0.164].动态触发条件下仿真结果如图2—6所示.

  • 图2 ηit)收敛曲线

  • Fig.2 Convergence curve of ηi (t)

  • 静态触发参数与动态触发参数设置相同,触发条件设置为θidiei2kh-σy^i2kh0,仿真结果如图7—11所示.

  • 图2和图7展示了ηi的收敛过程,可以看出动态触发条件下的结果与静态事件触发下的结果相同,4个智能体的成本函数最终都将稳定在优化值η1=η2=η3=η4=η*=19.320,仿真中的结果与利用式(6)计算得出结果相同.图3中的点表示本文所设计算法对应4个智能体的动态事件触发时刻,图8描述4个智能体的静态触发时刻,表2给出2种触发条件对应的触发次数,可以看出本文给出的方法触发频率更低,表明本文所设计的算法可有效降低带宽、减少通信负担.由图4和图9可以看出控制输入η˙it是分段函数,智能体仅在本身及邻居的事件触发时刻进行更新.图5和图10给出了mit)的收敛曲线,可以看出m1由初始值140稳定在99.575,m2由初始值110稳定在132.760,m3由初始值100稳定在91.055,m4由初始值90稳定在116.593.由图6和图11可以看出在趋于最优解的过程中,系统实时满足等式约束.

  • 图3 动态事件触发下每个智能体的事件触发时刻

  • Fig.3 Event triggering time of each agent under dynamic event triggered control

  • 图4 η˙it收敛曲线

  • Fig.4 Convergence curve of η˙i (t)

  • 图5 mi(t)收敛曲线

  • Fig.5 Convergence curve of mi (t)

  • 图6 i=14 mitM的轨迹

  • Fig.6 Trajectories of i=14 mi (t) and M

  • 图7 ηit)收敛曲线

  • Fig.7 Convergence curve of ηi (t)

  • 图8 静态事件触发下每个智能体的事件触发时刻

  • Fig.8 Event triggering time of each agent under static event triggered control

  • 图9 η˙it收敛曲线

  • Fig.9 Convergence curve of η˙i (t)

  • 图10 mit)收敛曲线

  • Fig.10 Convergence curve of mi (t)

  • 图11 i=14 mitM的轨迹

  • Fig.11 Trajectories of i=14 mi (t) and M

  • 表2 两种触发条件性能对比

  • Table2 Performance comparison of two triggering conditions

  • 4 总结

  • 本文研究了含有等式约束的二次凸优化问题,并针对这类问题设计了一种基于动态事件触发控制的分布式优化算法,该算法可以保证系统最终渐近收敛到最优解.在所设计的触发条件下,每个智能体仅需在自身触发时刻进行更新,不需要连续或周期性地更新控制信息,有效降低了智能体间通信频率以及控制器更新频率,并且通过引入周期采样控制,使触发时间存在下限值,避免了Zeno现象. Matlab仿真结果表明,与静态触发控制相比,所提出算法触发次数更少.

  • 参考文献

    • [1] 朱旭,张逊逊,闫茂德,等.基于一致性的无人机编队控制策略[J].计算机仿真,2016,33(8):30-34;ZHU Xu,ZHANG Xunxun,YAN Maode,et al.UAV formation control strategy based on consensus[J].Computer Simulation,2016,33(8):30-34

    • [2] 刘建刚,杨胜杰.具有容性负载的直流微电网系统分布式协同控制[J].自动化学报,2020,46(6):1283-1290;LIU Jiangang,YANG Shengjie.Distributed cooperative control of DC micro-grid systems with capacitive loads[J].Acta Automatica Sinica,2020,46(6):1283-1290

    • [3] 刘伟,李大卫,戴洪德,等.具有避障的机器人集群系统分布式编队控制[J].机械与电子,2021,39(7):44-48;LIU Wei,LI Dawei,DAI Hongde,et al.Distributed multi-robots system formation control with obstacle avoidance[J].Machinery & Electronics,2021,39(7):44-48

    • [4] 涂志亮,王强,沈毅.一种目标监测的移动传感器网络覆盖分布式优化算法[J].控制与决策,2012,27(9):1353-1358;TU Zhiliang,WANG Qiang,SHEN Yi.A distributed coverage optimization algorithm for target monitoring in mobile sensor network[J].Control and Decision,2012,27(9):1353-1358

    • [5] Kia S S,Cortés J,Martínez S.Distributed convex optimization via continuous-time coordination algorithms with discrete-time communication[J].Automatica,2015,55:254-264

    • [6] Xi C G,Khan U A.Distributed subgradient projection algorithm over directed graphs[J].IEEE Transactions on Automatic Control,2017,62(8):3986-3992

    • [7] Wen X N,Qin S T.A projection-based continuous-time algorithm for distributed optimization over multi-agent systems[J].Complex & Intelligent Systems,2021:1-11

    • [8] Yi P,Hong Y G,Liu F.Distributed gradient algorithm for constrained optimization with application to load sharing in power systems[J].Systems & Control Letters,2015,83:45-52

    • [9] 吴文,刘斌,姚靖,等.基于事件触发网络控制系统的分布式控制[J].湖南工业大学学报,2014,28(4):61-66;WU Wen,LIU Bin,YAO Jing,et al.Event-triggered based distributed control in networked control systems[J].Journal of Hunan University of Technology,2014,28(4):61-66

    • [10] Tabuada P.Event-triggered real-time scheduling of stabilizing control tasks[J].IEEE Transactions on Automatic Control,2007,52(9):1680-1685

    • [11] Dimarogonas D V,Frazzoli E,Johansson K H.Distributed event-triggered control for multi-agent systems[J].IEEE Transactions on Automatic Control,2012,57(5):1291-1297

    • [12] Liu Z X,Chen Z Q.Event-triggered average-consensus for multi-agent systems[C]//Proceedings of the 29th Chinese Control Conference(CCC’10),2010:4506-4511

    • [13] Xie D S,Xu S Y,Zhang B Y,et al.Consensus for multi-agent systems with distributed adaptive control and an event-triggered communication strategy[J].IET Control Theory & Applications,2016,10(13):1547-1555

    • [14] Fan Y,Feng G,Wang Y,et al.Distributed event-triggered control of multi-agent systems with combinational measurements[J].Automatica,2013,49(2):671-675

    • [15] Li H Q,Liu S,Soh Y C,et al.Achieving linear convergence for distributed optimization with zeno-like-free event-triggered communication scheme[C]//2017 29th Chinese Control and Decision Conference(CCDC).May 28-30,2017,Chongqing,China.IEEE,2017:6224-6229

    • [16] Du W,Yi X L,George J,et al.Distributed optimization with dynamic event-triggered mechanisms[C]//2018 IEEE Conference on Decision and Control.December 17-19,2018,Miami,FL,USA.IEEE,2018:969-974

  • 参考文献

    • [1] 朱旭,张逊逊,闫茂德,等.基于一致性的无人机编队控制策略[J].计算机仿真,2016,33(8):30-34;ZHU Xu,ZHANG Xunxun,YAN Maode,et al.UAV formation control strategy based on consensus[J].Computer Simulation,2016,33(8):30-34

    • [2] 刘建刚,杨胜杰.具有容性负载的直流微电网系统分布式协同控制[J].自动化学报,2020,46(6):1283-1290;LIU Jiangang,YANG Shengjie.Distributed cooperative control of DC micro-grid systems with capacitive loads[J].Acta Automatica Sinica,2020,46(6):1283-1290

    • [3] 刘伟,李大卫,戴洪德,等.具有避障的机器人集群系统分布式编队控制[J].机械与电子,2021,39(7):44-48;LIU Wei,LI Dawei,DAI Hongde,et al.Distributed multi-robots system formation control with obstacle avoidance[J].Machinery & Electronics,2021,39(7):44-48

    • [4] 涂志亮,王强,沈毅.一种目标监测的移动传感器网络覆盖分布式优化算法[J].控制与决策,2012,27(9):1353-1358;TU Zhiliang,WANG Qiang,SHEN Yi.A distributed coverage optimization algorithm for target monitoring in mobile sensor network[J].Control and Decision,2012,27(9):1353-1358

    • [5] Kia S S,Cortés J,Martínez S.Distributed convex optimization via continuous-time coordination algorithms with discrete-time communication[J].Automatica,2015,55:254-264

    • [6] Xi C G,Khan U A.Distributed subgradient projection algorithm over directed graphs[J].IEEE Transactions on Automatic Control,2017,62(8):3986-3992

    • [7] Wen X N,Qin S T.A projection-based continuous-time algorithm for distributed optimization over multi-agent systems[J].Complex & Intelligent Systems,2021:1-11

    • [8] Yi P,Hong Y G,Liu F.Distributed gradient algorithm for constrained optimization with application to load sharing in power systems[J].Systems & Control Letters,2015,83:45-52

    • [9] 吴文,刘斌,姚靖,等.基于事件触发网络控制系统的分布式控制[J].湖南工业大学学报,2014,28(4):61-66;WU Wen,LIU Bin,YAO Jing,et al.Event-triggered based distributed control in networked control systems[J].Journal of Hunan University of Technology,2014,28(4):61-66

    • [10] Tabuada P.Event-triggered real-time scheduling of stabilizing control tasks[J].IEEE Transactions on Automatic Control,2007,52(9):1680-1685

    • [11] Dimarogonas D V,Frazzoli E,Johansson K H.Distributed event-triggered control for multi-agent systems[J].IEEE Transactions on Automatic Control,2012,57(5):1291-1297

    • [12] Liu Z X,Chen Z Q.Event-triggered average-consensus for multi-agent systems[C]//Proceedings of the 29th Chinese Control Conference(CCC’10),2010:4506-4511

    • [13] Xie D S,Xu S Y,Zhang B Y,et al.Consensus for multi-agent systems with distributed adaptive control and an event-triggered communication strategy[J].IET Control Theory & Applications,2016,10(13):1547-1555

    • [14] Fan Y,Feng G,Wang Y,et al.Distributed event-triggered control of multi-agent systems with combinational measurements[J].Automatica,2013,49(2):671-675

    • [15] Li H Q,Liu S,Soh Y C,et al.Achieving linear convergence for distributed optimization with zeno-like-free event-triggered communication scheme[C]//2017 29th Chinese Control and Decision Conference(CCDC).May 28-30,2017,Chongqing,China.IEEE,2017:6224-6229

    • [16] Du W,Yi X L,George J,et al.Distributed optimization with dynamic event-triggered mechanisms[C]//2018 IEEE Conference on Decision and Control.December 17-19,2018,Miami,FL,USA.IEEE,2018:969-974

  • 地址:江苏省南京市宁六路219号    邮编:210044

    联系电话:025-58731025    E-mail:nxdxb@nuist.edu.cn

    南京信息工程大学学报 ® 2024 版权所有  技术支持:北京勤云科技发展有限公司