具有切换拓扑的多智能体系统包围控制问题研究
发布时间：2018-07-23 12:46
【摘要】：由多个简单个体组成的群体称为多智能体系统,其协作控制主要是指个体通过交换局部信息而完成给定的群体任务,如多飞行器编队、多电机协同、多机器人协作搬运等都属于典型的协作控制问题。由于此类系统广泛存在于自然界和现实生活中,因此吸引了控制领域、物理等众多领域学者展开研究,并取得了丰硕成果。其中,包围控制是该领域中的一个基本问题,其目的是通过设计合适的基于个体的控制协议,使得追随者收敛到由多个领导者组成的凸包内。本文针对个体之间相邻关系不断切换条件下的包围控制问题展开研究,主要内容如下:1.研究了信息交换拓扑随机切换的一阶多智能体系统包围控制问题。考虑由多个静态领导者和追随者组成的一阶连续多智能体系统,基于所设计的控制协议,证得当有向信息交换联合图全局可达时,系统可解包围控制问题,仿真结果验证了结果的有效性。2.研究了信息交换拓扑随机切换的二阶多智能体系统包围控制问题。考虑由多个动态领导者和追随者组成的二阶离散多智能体系统,基于所设计的控制协议,通过构造一个均方误差函数,对系统进行等价变换,将二阶包围控制问题转化为降阶的系统稳定性问题,并得到了误差系统稳定的充分条件。即在此条件下,系统可解包围控制问题,仿真结果验证了结果的有效性。3.在切换通信拓扑条件下,研究了一类多智能体系统的环绕控制问题。在二维平面上,考虑由一个静态领导者和多个追随者组成的多智能体系统,控制目标是所有追随者最终收敛到以领导者为圆心的圆周上,并均匀分布在该圆周上做环绕运动。为此,首先给出了一个追随者估计静态领导者位置的分布式控制协议,并得出了所有追随者收敛到指定圆周上的充分条件。其次,设计了追随者在圆周上运动的分布式控制协议,使系统在满足前述充分条件时,所有追随者均匀分布在圆周上并做环绕运动。仿真结果验证了所提方法的有效性。
[Abstract]:A group composed of many simple individuals is called a multi-agent system. Its cooperative control mainly refers to the individual accomplishing a given group task by exchanging local information, such as multi-aircraft formation and multi-motor cooperation. Multi-robot cooperative handling is a typical cooperative control problem. Because this kind of system exists widely in nature and real life, it attracts a lot of researchers in the field of control, physics and so on, and has made a lot of achievements. Encircling control is a basic problem in this field. Its purpose is to make followers converge to convex hull composed of multiple leaders by designing appropriate individual-based control protocols. In this paper, we study the bounding control problem under the condition of continuous switching between individuals, the main contents are as follows: 1. The encircling control problem of a first order multi-agent system with random switching of information exchange topology is studied. Considering the first-order continuous multi-agent system composed of many static leaders and followers, based on the designed control protocol, it is proved that when the directed information exchange joint graph is globally reachable, the system can solve the problem of surround control. Simulation results verify the effectiveness of the results. 2. In this paper, the encircling control problem of second order multi-agent systems with information exchange topology switching is studied. The second order discrete multi-agent system composed of multiple dynamic leaders and followers is considered. Based on the designed control protocol, a mean square error function is constructed to transform the system equivalent. The second order bounding control problem is transformed into the stability problem of reduced order systems, and the sufficient conditions for the stability of error systems are obtained. That is to say, under this condition, the system can be surrounded by the solvable control problem. The simulation results verify the effectiveness of the results. 3. Under the condition of switched communication topology, the circumferential control problem of a class of multi-agent systems is studied. In a two-dimensional plane, a multi-agent system consisting of a static leader and a plurality of followers is considered. The control goal is that all the followers converge to the circle centered on the leader and move around the circle uniformly. To solve this problem, a distributed control protocol is proposed to estimate the position of the static leader, and a sufficient condition for all followers to converge to the specified circle is obtained. Secondly, a distributed control protocol for the movement of followers on the circle is designed, so that when the system meets the above sufficient conditions, all the followers are evenly distributed on the circle and move around. Simulation results show the effectiveness of the proposed method.
【学位授予单位】：兰州交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18;TP13
本文编号：2139457
[Abstract]:A group composed of many simple individuals is called a multi-agent system. Its cooperative control mainly refers to the individual accomplishing a given group task by exchanging local information, such as multi-aircraft formation and multi-motor cooperation. Multi-robot cooperative handling is a typical cooperative control problem. Because this kind of system exists widely in nature and real life, it attracts a lot of researchers in the field of control, physics and so on, and has made a lot of achievements. Encircling control is a basic problem in this field. Its purpose is to make followers converge to convex hull composed of multiple leaders by designing appropriate individual-based control protocols. In this paper, we study the bounding control problem under the condition of continuous switching between individuals, the main contents are as follows: 1. The encircling control problem of a first order multi-agent system with random switching of information exchange topology is studied. Considering the first-order continuous multi-agent system composed of many static leaders and followers, based on the designed control protocol, it is proved that when the directed information exchange joint graph is globally reachable, the system can solve the problem of surround control. Simulation results verify the effectiveness of the results. 2. In this paper, the encircling control problem of second order multi-agent systems with information exchange topology switching is studied. The second order discrete multi-agent system composed of multiple dynamic leaders and followers is considered. Based on the designed control protocol, a mean square error function is constructed to transform the system equivalent. The second order bounding control problem is transformed into the stability problem of reduced order systems, and the sufficient conditions for the stability of error systems are obtained. That is to say, under this condition, the system can be surrounded by the solvable control problem. The simulation results verify the effectiveness of the results. 3. Under the condition of switched communication topology, the circumferential control problem of a class of multi-agent systems is studied. In a two-dimensional plane, a multi-agent system consisting of a static leader and a plurality of followers is considered. The control goal is that all the followers converge to the circle centered on the leader and move around the circle uniformly. To solve this problem, a distributed control protocol is proposed to estimate the position of the static leader, and a sufficient condition for all followers to converge to the specified circle is obtained. Secondly, a distributed control protocol for the movement of followers on the circle is designed, so that when the system meets the above sufficient conditions, all the followers are evenly distributed on the circle and move around. Simulation results show the effectiveness of the proposed method.
【学位授予单位】：兰州交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18;TP13
【参考文献】
相关期刊论文 前5条
1 段敏;高辉;宋永端;;智能群体环绕运动控制[J];物理学报;2014年14期
2 刘帅;谢立华;张焕水;;带噪声多自主体的均方包容控制(英文)[J];自动化学报;2013年11期
3 张文广;刘吉臻;曾德良;胡勇;;Consensus of second-order multi-agent systems with nonuniform time delays[J];Chinese Physics B;2013年05期
4 王晓丽;洪奕光;;多智能体系统分布式控制的研究新进展[J];复杂系统与复杂性科学;2010年Z1期
5 胡江平;袁海文;;Collective coordination of multi-agent systems guided by multiple leaders[J];Chinese Physics B;2009年09期
相关博士学位论文 前2条
1 董立静;变拓扑结构下多智能体系统的追踪问题[D];北京理工大学;2016年
2 唐朝君;多智能体系统一致性问题与包含控制问题研究[D];电子科技大学;2012年
，本文编号：2139457
本文链接：https://www.wllwen.com/kejilunwen/zidonghuakongzhilunwen/2139457.html
