异质多智能体系统的分布式协调问题研究

In the practical production and life, the dynamics of the agents coupled with others are different because of various restrictions or the common goals with mixed agents. However, previous researches of distributed coordination mainly facus on the homogeneous multi-agent systems. In this project, the distributed coordination of heterogeneous multi-agent systems will be considered, in which the agents are first-order and second-order integrator agents. The first objective is to design the distributed protocols for heterogeneous multi-agent systems with time-delay, quantization and loss in communication. By using the Lyapunov theory, algebraic graph theory and matrix theory, some sufficient conditions for consensus of heterogeneous multi-agent systems will be established. The second purpose is to study the flocking of heterogeneous multi-agent systems with state-dependent connectivity. Through designing the communication rate functions and distributed protocols, some sufficient conditions for solving the flocking of heterogeneous multi-agent systems will be obtained by applying the Lyapunov theory and Lasalle invariance principle. The third objective is to study the containment control, formation control and coverage control of heterogeneous multi-agent systems. By using the methods which are employed to study the distributed coordination of homogeneous multi-agent systems and consensus of heterogeneous multi-agent systems, some sufficient conditions or necessary and sufficient conditions will be obtained.

现实生产和生活中，许多任务需要不同动力学结构的智能体共同协作完成，而先前的研究主要集中在具有相同动力学结构的齐次多智能体系统中，本项目研究由一阶和二阶积分器构成的异质多智能体系统的分布式协调问题。(1)针对通信中信息延迟、量化和丢失的情况，设计分布式协议，利用Lyapunov理论、代数图论和矩阵理论，得到异质多智能体系统达到一致的充分条件；(2)研究通信与状态相关的异质多智能体系统的群集行为。通过设计通信率函数和分布式协议，利用Lyapunov理论和Lasalle不变集原理，得到解决异质多智能体系统群集行为的充分条件；(3)针对不同情形下，设计有效的分布式协议，利用先前研究同类齐次多智能体系统的分布式协调问题和异质多智能体系统一致性问题的方法，得到不同通信拓扑下解决异质多智能体系统解决包围控制、编队控制和覆盖控制的充分/充要条件。

现实生产和生活中，许多任务需要不同动力学结构的智能体共同协作完成。本项目研究了异质多智能体系统的分布式协调问题，主要研究内容和结果如下：（1）根据异质多智能体系统的特点，设计了一类新奇的群一致性协议；设计了异质多智能体系统在信息量化下的协议，并得到其达到一致性的基本判据准则；（2）研究了异质多智能体系统的包围控制问题，得到系统解决包围控制的充要判据；（3）研究了通信与状态相关的多智能体系统的群集和一致性问题，得到系统解决协同控制时通信函数满足的上下界；（4）进一步深入研究了多智能体系统的量化一致性和有限时间一致性，得到系统解决一致性的充分判据；（5）提出并研究了切换多智能体系统的分布式协调问题，得到解决切换多智能体系统分布式协调的充分/充要判据；（6）通过设计指标函数，研究了多智能体系统解决分布式协调问题时最优/次优拓扑，给出拓扑选择的充分/必要条件。. 研究成果建立了解决齐次多智能体系统和异质多智能体系统分布式协调问题的统一框架，揭示了一阶多智能体系统与二阶多智能体系统分布式协调的内在关联。采用微分包含理论、最优化方法、博弈论等工具分析多智能体系统的分布式协调问题，对已有的研究工具进行了必要的拓展，得到一些解决分布式协调问题的新工具和新结果。根据现实工程需求，提出了一类多智能体系统，并分析其分布式协调问题等等。