输入约束下的多智能体系统完全分布式协调控制研究

Coorperative control of multi-agent systems is a hot interdisciplinary research area for mathematics, control science, physics, biology and so on. It has an engineering significance due to its potential use in multi-sensor networks, formations of multi-UAVs (unmanned aerial vehicles), industrial cooperative transportaion and construction. For a class of multi-agent systems with input contraints, most researches assumed that the global network information is avaliable and the network topology is fixed. This limits the application range for multi-agent systems to some extent. To tackle these problems, this project will investigate the issues including modelling, dynamic analysis, fully distributed controller designing for the multi-aget systems with input constraints under the general jointly connected topologies, based on algebraic graph theory, matrix theory, adaptive control technics, Laypunov stability theory and other nonlinear stability analytical tools. It will effectively reduce the controller cost and enhance the calculation efficiency for the system. Research achievements will boost the development of mulit-agent systems, and also provide the theoretical and pratical basis for the use in industrial production, national defense and military.

多智能体系统的协调控制是数学、控制科学、物理、生物等学科的交叉前沿研究领域。它在多传感器网络、多无人机编队、工业协同运输建造等领域具有重要的实际工程意义。目前国内外关于输入约束下的多智能体系统研究大多假定网络全局信息是可利用的并且网络拓扑是固定不变的，这一定程度上限制了多智能体系统的工程应用范围。为了克服这些问题，本项目拟基于代数图理论、矩阵理论、自适应控制技术、Laypunov稳定性理论及其他非线性稳定性分析工具，研究一般的联合连通切换拓扑条件下具有输入约束的多智能体系统的建模、动力学分析、完全分布式控制器设计等问题。通过本项目研究，将有效地降低控制输入成本，提高系统的运算效率。研究成果将推动多智能体系统领域的发展，并为工业生产和国防军事中的应用提供理论与应用依据。

多智能体系统的协调控制是数学、控制科学、物理、生物等学科的交叉前沿研究领域。它在多传感器网络、多无人机编队、工业协同运输建造等领域具有重要的实际工程意义。本项目首先针对一类具有输入约束和未知控制系数的二阶多智能体系统，给出了分布式一致性控制设计算法，并利用自适应控制技术、矩阵理论、稳定性理论、代数图理论证明了系统的渐近收敛。另外，本项目还研究了具有未知非线性动力学行为和未知控制系数下的一阶和二阶多智能体系统的一致性问题，并利用自适应控制技术、牵制控制技术、Nussbaum增益技术、代数图理论以及Barbalat引理，理论证明了系统的渐近一致稳定。本项目对于输入约束下的多智能体系统一致性问题提供了初步的技术方法，该工作具有重要的理论意义和应用价值。