多智能体系统的量化一致性与包围控制

This project will consider the consensus and containment control problems for multi-agent systems in the aspects of node dynamics and network effects, and present a unified framework for both discrete-time and continuous-time multi-agent systems from different perspectives. The specific objectives of the project include: (1) The first objective is to investigate the quantized consensus of switched multi-agent systems, which are composed of continuous-time and discrete-time subsystems. Pinning and impulsive control strategies will also be proposed to solve the quantized consensus problem in switched multi-agent systems with a leader; (2) The second objective is to the study the consensus of diserete-time multi-agent system described by Delta operator in the presence of quantizers and faulty agents. (3) The third objective is to get some criteria for solving the containment control problem of the above-mentioned two classes of multi-agent systems. This project covers an integration of multi-disciplinary research areas, including algebraic graph theory, stability theory, differential inclusion theory, impulsive control theory, Delta operator theory, stochastic analysis tools and wireless communication. This project will enrich and develop the theory in networked multi-agent systems by introducing new models and designing new algorithm.

本项目从节点动力学和网络效应方面研究多智能体系统的一致性和包围控制问题，建立研究连续和离散时间多智能体系统一致性与包围控制问题的统一框架。主要研究内容如下：(1) 针对由连续和离散时间子系统构成的切换多智能体系统，研究其量化一致性问题，并基于牵引和脉冲策略，解决具有单领航者的量化一致性问题; (2) 基于Delta算子建立几类多智能体系统的离散化模型，研究系统在量化通讯和存在故障节点情况下的一致性问题; (3) 研究上述多智能体系统的包围控制问题，建立包围控制可解的基本判据。本项目融合代数图论、稳定性理论、微分包含理论、脉冲控制理论、Delta算子理论、随机分析工具和无线通信等多学科交叉技术与知识，探求新模型，设计新算法和提供新的研究方案，丰富和发展网络化多智能体系统的分析与控制。

在过去三年中，项目申请人与项目组成员对本项目书拟定的主要研究内容进行了深入细致的研究，主要成果包括：针对对抗网络下的多智能体系统，设计了基于对数量化信息的分布式协议，得到系统解决二分一致性的图判据；针对存在恶意个体的异质多智能体系统，设计了安全一致性算法，结合网络鲁棒性概念，得到系统解决安全一致性的图判据；提出了几类由连续时间和离散时间个体混杂的多智能体系统，设计分布式协议，得到系统解决一致性的充分/充要判据；针对异质多智能体系统，设计了有/无速度信息的非线性协议解决了系统的赢者通吃问题，并揭示了最终赢者与外输入间的关系；给出了一类非线性正系统固定时间稳定的充分判据，扩广了多智能体系统固定时间一致性的分析工具。本项目以理论研究为主，丰富和发展了网络化多智能体系统的分析与控制，包括拓广分析工具，丰富研究内容等。