具有动态不确定性的下三角多智能体系统分布式自适应协同控制

This project is concerned with the distributed cooperative control problem for multi-agent systems in lower-triangular form with uncertain dynamics. Considering complex network communication environment, we will be dedicated to developing cooperative control approaches for a class of typical nonlinear multi-agent systems combining with event-triggered mechanism and nonlinear adaptive control techniques. Firstly, state feedback/output feedback distributed cooperative control continuous strategies will be proposed via measurable/immeasurable local information of the agents. In additional, distributed information exchanging mode will be established based on event-triggered scheme to reduce network communication pressure. Further, we shall discuss the relationship between parameters of the controller and the proposed trigger, and then a distributed cooperative control law will be presented based on the designed event-triggered scheme. Last but not the least, a distributed fault detection and fault tolerant controller will be under discussion and appropriate choice of performance index, together with the analysis of multi objective optimization methods and more concise control laws, will be studied. In this project, a novel theoretical framework for uncertain multi-agent systems in lower-triangular form will be established. It is expected that this work could provide theoretical foundation for coordinated mobiles, ocean vehicles formation etc.

项目主要研究具有动态不确定性的下三角多智能体系统的协同控制问题。综合考虑复杂网络通信环境，结合事件触发机制、非线性自适应控制技术，探索解决一大类典型的非线性多智能体系统的协同控制方法。首先，在智能体状态可测/不可测的情况下，利用局部信息，讨论纯连续策略下状态反馈/输出反馈分布式自适应协同控制问题；其次，为减轻网络数据传输压力，建立基于事件触发机制的分布式信息交互方式，讨论控制器参数和触发条件参数关系，设计触发机制下的分布式自适应协同控制律；最后，为保证网络通信可靠性，讨论分布式故障诊断与容错控制器的设计，选择合适的性能指标，解决多目标优化问题，同时尽可能使得控制律结构简捷。基于上述工作，将初步形成不确定下三角多智能体系统控制方法体系，为机器人协同工作、远洋舰船编队等工程问题提供理论支撑。

本课题以不确定下三角非线性多智能体系统为研究对象，基于backstepping、自抗扰技术、图论、神经网络逼近理论、触发机制，考虑状态反馈/输出反馈、含时滞项和输入饱和限制、执行机构含故障等情况，面向一大类典型的多智能体系统输出信号协同一致问题进行了系统的研究。本项目的主要研究内容和结果有：1) 针对一类非仿射纯反馈非线性多智能体系统，设计自适应状态反馈和输出反馈分布式协同控制策略；考虑含时滞项和输入限幅的一类非仿射纯反馈多智能体系统，设计自适应分布式输出一致控制策略；2) 结合自抗扰技术，研究非仿射非线性多智能体系统的输出一致控制策略和输出反馈包含控制策略；3) 基于动态面控制技术，研究一类不确定MIMO下三角时滞非线性系统的自适应状态反馈控制和输出反馈跟踪控制方法；4) 研究一类非线性多智能体系统状态反馈和输出反馈事件触发/自触发协同控制方法，结合动态规划方法设计近似最优控制律；5) 针对含执行器故障的一类非仿射纯反馈多智能体系统，考虑失效和偏置等故障，研究自适应分布式输出一致容错控制策略；6) 搭建网络化实验验证仿真平台，并开展实验研究。上述工作为工程应用提供理论支撑。