几类多智能体系统的量化一致性和包围控制

发布时间：2018-08-28 12:45

【摘要】：由于多智能体系统的分布式协调控制在许多领域都有潜在的应用,它已经成为复杂系统研究的一个热门课题,受到了不同领域研究者的广泛关注。在多智能体系统中,个体通过互相通信,彼此协作,共同完成某个复杂任务。而随着数字系统在信息传输和处理中越来越广泛的应用,研究量化误差对系统性能的影响具有重要的实际意义。本文研究多智能体系统分布式协调控制的两类问题：量化一致性问题和包围控制问题。主要研究内容如下：1.研究了具有非线性动力学的多智能体系统的量化一致性问题。对一阶和二阶非线性多智能体系统,设计了基于邻居智能体之间相对状态量化值的分布式协议。在连续时间协议下,综合利用非光滑分析工具,得到了系统解决一致性的充分条件。在脉冲协议下,证明了在一定的量化器精度下可以通过选择合适的控制增益和脉冲间隔使得所有智能体的状态达到一致。2.研究了由一阶和二阶积分器智能体构成的异质多智能体系统的量化一致性问题。设计了基于邻居之间相对位置量化值和绝对速度量化值的分布式协议。应用图论和拉萨尔不变集原理,证明了在使用对数量化器时,系统可以渐近达到静态一致。而在使用均匀量化器时,通过选择合适的控制增益,二阶智能体的速度收敛到零,而智能体之间的相对位置将收敛到一个有界区域。3.研究了切换多智能体系统的量化一致性问题。本文所指的切换多智能体系统由连续时间和离散时间子系统所构成。对一阶和二阶切换多智能体系统,设计了基于邻居智能体之间相对状态量化值的分布式协议。对一阶切换多智能体系统,得到了系统在任意切换下达到一致的充分条件。对二阶切换多智能体系统,证明了对任意给定脉冲间隔,可以通过选取适当的控制增益和充分小的量化器精度,使得所有智能体的状态达到一致。4.研究了切换多智能体系统的包围控制问题。为每个智能体设计了基于邻居之间相对状态值的分布式协议。应用图论、稳定性理论等工具,分别对具有静态和动态领航者的切换多智能体系统,得到了系统在任意切换下达到包围控制的充要条件。
[Abstract]:As the distributed coordinated control of multi-agent systems has potential applications in many fields, it has become a hot topic in the research of complex systems, and has been widely concerned by researchers in different fields. In a multi-agent system, individuals work together to complete a complex task by communicating and cooperating with each other. With the more and more extensive application of digital system in information transmission and processing, it is of great practical significance to study the effect of quantization error on system performance. In this paper, we study two kinds of distributed coordinated control problems in multi-agent systems: quantization consistency problem and bounding control problem. The main research contents are as follows: 1. The quantization consistency problem of multi-agent systems with nonlinear dynamics is studied. For first and second order nonlinear multi-agent systems, a distributed protocol based on relative state quantization between neighbor agents is designed. Under the continuous time protocol, a sufficient condition for the system to solve the consistency is obtained by using the non-smooth analysis tool. Under the pulse protocol, it is proved that the state of all agents can be consistent by selecting the appropriate control gain and pulse interval under certain quantizer precision. The quantization consistency problem of heterogeneous multi-agent systems composed of first and second order integrator agents is studied. A distributed protocol based on relative position quantization and absolute velocity quantization between neighbors is designed. By using graph theory and Lassar invariant set principle, it is proved that the system can asymptotically achieve static consistency when using logarithmic quantizer. When the uniform quantizer is used, the velocity of the second order agent converges to zero by selecting the appropriate control gain, and the relative position of the agents converges to a bounded region. 3. The quantization consistency problem of switched multi-agent system is studied. The switched multi-agent system in this paper consists of continuous time subsystem and discrete time subsystem. For first-order and second-order switched multi-agent systems, a distributed protocol based on relative state quantization between neighbor agents is designed. For first-order switched multi-agent systems, sufficient conditions for the system to be consistent in arbitrary handoff are obtained. For a second-order switched multi-agent system, it is proved that for any given pulse interval, the state of all agents can be consistent with .4by selecting appropriate control gain and sufficiently small quantizer accuracy. The bounding control problem of switched multi-agent system is studied. A distributed protocol based on relative state value between neighbors is designed for each agent. By means of graph theory and stability theory, the necessary and sufficient conditions of bounding control are obtained for switched multi-agent systems with static and dynamic navigators respectively.
【学位授予单位】：西安电子科技大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TP18

【相似文献】