摘要当今世界，随着计算机技术，人工智能技术的提高，随着网络的出现和飞速发展，集中控制逐渐不能适应高速发展的科学技术，因而集中式计算和分布式处理技术诞生了。对于大规模的网络系统的控制，分布式控制更为灵活方便，因而，分布式控制取得了飞速的发展。33393
最近几年，多智能体系统（Multi-agent Systems）分布式协调控制问题引起了物理学，生物学，控制类科学，计算机科学等领域的科学研究者的广泛关注。并且在无人驾驶飞机的编队控制，多智能体群集运动，分布式传感器网络，交通拥堵控制，人造卫星位姿调整等有了广泛的应用。在多智能体的协调控制当中，其基本问题包括了编队控制的问题，聚集控制的问题，群集运动，以及一致性问题。尤为重要的就是一致性问题。一致性问题，更是分布式计算理论的基础。所谓的一致性也就是随之时间的变换，多智能体系统中所有主体的状态最后趋向于一致。
本文首先介绍了多智能体系统分布式协调控制的研究背景以及其研究意义，并阐述了其存在的的基本问题，同时对相关预备知识作了简要说明。其次，研究了根据群集运动Leader-follow模式下的智能群体协同控制。给出分布式控制器的设计，跟随者对动态领导者的跟踪误差一致最终有界,且最终收敛到原点可调区域内。并结合Lyapunov稳定性分析方法，给出了参数选择的范围。最后通过计算机仿真验证了结论的准确性。
关键词 多智能体系统 协调控制 一致性问题 群集运动
毕业论文设计说明书外文摘要
Title Nonlinear Multi-agent Systems Distributed Cooperative Control
Abstract
In recent years,With the development of computer technology and artificial intelligence technology and the appearance of network,the centralized control gradually can not meet the rapid development of science and technology.And the centralized computing and distributed processing technology emerges.In the control of large-scale network systems,the distributed control is more flexible and convenient.Therefore,Distributed control has achieved rapid development.
In recent years,the distributed cooperative control problem of Multi-agents Systems has attracted the researchers of Physics, biology, control and computer science.Such as unmanned air vehicles formation,swarming,distributed sensor network,traffic congestion control and attitude alignment of cluters of satellite.The basic problems of Multi-agent cooperative control include formation control problem,swarming,aggregation control problems and consensus problem. Consensus problem is the basic theory of distributed computing.The consensus means that the status of all the subjects convergence as time increases.
This article introduced multi-agent systems distributed cooperative control’s research background and significance and the basic problems,firstly.Then,we study the group cooperative control under the Leader-follow state,according to the swarming and flocking.then,we design the distributed controller,find that the consensus errors are cooperatively semiglobally uniformly ultimately bounded and converge to an adjustable neighborhood of the origin. And we combine Lyapunov stability to analysis,then give the range of preferences.Finally,We verified the accuracy of the conclusions by the computer simulation.
Keywords Multi-agent Systems Cooperative control Consensus problem Swarming and flocking
目 次
1 绪论 1
1.1 研究背景与意义 1
1.2 国内外研究现状 2
1.3 多智能体系统协调控制的几个基本问题 2
1.31 编队控制问题 3
1.32 聚集控制问题 3
1.33 群集运动 4
1.34 一致性问题 5 非线性多智能体系统分布式协调控制:http://www.751com.cn/zidonghua/lunwen_30439.html