基于角色切换的混杂多智能体网络的协调控制

The role switching and hybrid feature are very common in biological and artificial networks, and are the inherent characteristics of multi-agent networks. Role switching refers to the role of the network node can dynamically switch with mission requirements. Hybrid feature is the mix of the network parameters jump, heterogeneous, topology switching. These features will give great challenge to network dynamics analysis and coordination control and is a difficult problem in the field of the network control. In this project, based on the role evolution, the coordination control of hybrid multi-agent networks will be studied. The main content include: 1) Based on the mixture of node and the topological structure in network, the dynamic models of hybrid multi-agent networks will be established; 2) With the role switching of hybrid multi-agent network, this project study the dynamical behavior and coordination control of hybrid multi-agent networks; 3) According to the uncertainties of complex network environment and the role switching, the coordination control of hybrid multi-agent networks will be investigated; 4) Utilizing the simulation software and experiment platform of multiple robots, verify the effectiveness of the proposed theoreticla results and improve the algorithms. Based on the nature of role switching of real networks, this project studies the dynamical behavior and cooperative control problem of hybrid multi-agent networks, aiming to enrich and develop new theories and technologies for networked control systems, and provide some theoretical guidance for the design of real networks.

角色切换与混杂特性在生物和人工网络中普遍存在，是多智能体网络的固有特征。角色切换是指网络节点的角色会随着任务需求而动态切换；混杂特性是指网络中参数跳变、异质异构、拓扑切换等多因素混杂。上述特性会给网络动力学分析与协调控制带来极大挑战，是目前网络控制领域的重难点问题。本项目研究基于角色切换的混杂多智能体网络的协调控制问题，主要内容包括：1）基于网络中节点和拓扑结构的混杂特性，构建混杂多智能体网络的动力学模型；2）针对具有角色切换的混杂多智能体网络，研究其动力学分析与协调控制；3）具有不确定性及角色切换的混杂多智能体网络的动力学分析与控制；4）构建软硬件仿真实验平台以验证和改进得到的理论结果。本项目从实际网络的角色切换与混杂特性等固有特征出发，研究具有普适性的混杂多智能体网络的动力学行为与协调控制问题，旨在丰富和发展网络控制的新理论与新技术，为实际网络的开发和设计提供理论指导。

在过去三年中，项目负责人和其团队成员围绕申请书拟定的研究计划，开展了混杂多智能体网络的协调控制研究。从多智能体网络的协调控制理论角度，根据实际网络的混杂特性，智能体在网络中所处的不同角色，建立新的多约束多智能体网络模型，设计新的协调控制算法。研究了带分层领航者的多智能体混杂协调控制、分层多智能体系统的一致性控制、二阶多智能体系统的量化包含控制、二阶非线性多智能体系统的编队控制等情况。同时，还考虑了群集智能优化算法在移动机器人路径规划中的应用、双边移动机械臂遥操作系统的同步控制等问题。这些研究一方面在理论上拓展了多智能体系统的协调控制理论；另一方面，相应结果也将为多智能体系统的动力学分析与设计提供有力的理论指导。总的来说，在基金项目的支持下，我们取得了一些成果，为下一步的研究计划奠定了良好的基础。