基于事件触发策略的非线性多智能体的一致性研究

发布时间：2018-04-22 07:21

本文选题：时间延时 + 事件触发控制　；参考：《西南大学》2016年硕士论文

【摘要】：近年来,多智能体系统已经引起了来自于不同学科研究人员的广泛关注。这主要是因为它在很多方面有着广泛的应用前景,其研究成果被广泛的应用于机器人合作控制、车辆交通管理、电子商务、无人飞行器的协调控制、编队控制、群集、网络资源的分配控制。在多智能体系统中,一致性意味着系统中所有智能体的状态通过智能体之间的信息交换与融合,最终达到趋同。而一致性协议是智能体间相互作用、信息传递的规则,在这个过程中,智能体之间的信息交换协议起到了决定性的作用。根据多智能体系统所处的客观环境,人们往往设计出不同的信息交换协议,正是在这样的应用背景下,一致性问题成为最具有研究价值的问题之一。本文以多智能体系统为研究背景,重点分析了基于事件触发控制的非线性多智能体领航跟随一致性问题。文章的主要内容和研究成果总结如下:(1)简要介绍了多智能体系统的研究背景与研究意义以及本论文所做的工作。(2)研究了非线性多智能体系统存在输入时延的领航-跟随一致性问题。首先考虑网络拓扑结构为固定不变的,设计了相应的事件触发控制协议。假设系统的通信拓扑包含一棵有向生成树,运用代数图论、矩阵理论和Lyapunov方法,给出了非线性多智能体达到领航跟随一致的充分条件。接着,考虑了切换拓扑结构的网络情形,其相应的理论结果和固定拓扑结构类似。最后的仿真结果验证了理论的可行性和方法的有效性。(3)研究了基于事件触发控制的非线性多智能体半全局领航-跟随一致性问题。首先,为了克服传统的均匀采样控制的缺点,进一步减少控制器跟新的频率,减少不必要的计算和通讯资源的浪费,将事件触发控制协议引入到多智能体系统一致性问题研究中。本文提出了一种新的事件触发控制策略,它可以避免连续的通信。根据设定的控制器,得到了领航-跟随一致的充分条件。其次,为了进一步减少控制的消耗,本论文采用牵制控制策略,即只有控制器应用控制动作智能体的很少一部分。然而,大多数现有工作都基于不同智能体之间连续时间的通信,所设计的控制策略要求的实时更新,这就要求智能网络体必须配备高性能处理器和高速通信信道。最后,通过Lyapunov稳定性理论方法,本文找到时间延时非线性多智能体在事件触发机制下,达到一致所需要的条件。
[Abstract]:In recent years, multi-agent systems have attracted wide attention from researchers from different disciplines. This is mainly because it has a wide application prospect in many aspects, and its research results are widely used in robot cooperative control, vehicle traffic management, electronic commerce, coordination control of unmanned aerial vehicles, formation control, cluster. Network resource allocation control. In a multi-agent system, consistency means that the state of all agents in the system is exchanged and fused by the agents, and eventually converges. Conformance protocol is the rule of interaction and information transfer between agents. In this process, the information exchange protocol between agents plays a decisive role. According to the objective environment of multi-agent system, people often design different information exchange protocols. In such an application background, the consistency problem has become one of the most valuable problems. In this paper, based on the multi-agent system, the problem of nonlinear multi-agent pilotage consistency based on event-triggered control is analyzed. The main contents and results of this paper are summarized as follows: (1) A brief introduction to the research background and significance of multi-agent system and the work done in this thesis. -following consistency. Considering that the network topology is fixed and invariant, the corresponding event trigger control protocol is designed. Assuming that the communication topology of the system consists of a directed spanning tree, the sufficient conditions for the nonlinear multi-agent to achieve the homogeneity of pilotage following are given by using algebraic graph theory, matrix theory and Lyapunov method. Then, the network with switched topology is considered, and the corresponding theoretical results are similar to those of fixed topology. Finally, the simulation results verify the feasibility of the theory and the effectiveness of the method. We study the semi-global piloting-follow consistency problem of nonlinear multi-agent based on event-triggered control. First of all, in order to overcome the shortcomings of traditional uniform sampling control, further reduce the controller and new frequency, reduce unnecessary computing and communication resources waste, The event trigger control protocol is introduced into the research of multi-agent system consistency. In this paper, a new event trigger control strategy is proposed, which can avoid continuous communication. According to the set controller, a sufficient condition of piloting-follow consistency is obtained. Secondly, in order to further reduce the consumption of control, this paper adopts the containment control strategy, that is, only a small part of the controller application control action agent. However, most of the existing work is based on continuous time communication between different agents, and the designed control strategy requires real-time updating, which requires the intelligent network body to be equipped with high performance processors and high speed communication channels. Finally, through the Lyapunov stability theory method, we find the condition that the time-delay nonlinear multi-agent can reach the consistent requirement under the event triggering mechanism.
【学位授予单位】：西南大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP18

【相似文献】