随机时延下多输入多输出多智能体系统事件触发双向编队  

Event-triggered bipartite formation for multi-input multi-output multi-agent systems with random delays

作　　者：赵华荣 彭力 吴治海 谢林柏 于洪年 ZHAO Hua-rong;PENG Li;WU Zhi-hai;XIE Lin-bo;YU Hong-nian(The Engineering Research Center of Internet of Things Applications of Ministry of Education,Jiangnan University,Wuxi 214122,China;The School of Computing,Engineering and the Built Environment,Edinburgh Napier University,Edinburgh,EH105DT,UK)

机构地区：[1]江南大学物联网应用技术教育部工程中心,江苏无锡214122 [2]爱丁堡龙比亚大学计算机、工程与建筑环境学院,爱丁堡EH105DT

出　　处：《控制与决策》2024年第4期1251-1259,共9页Control and Decision

基　　金：国家自然科学基金项目(61873112,61802107,61876073);江苏省高等学校自然科学研究面上项目(18KJB413009)。

摘　　要：针对动力学模型未知的多输入多输出非线性离散时间多智能体系统的随机时延问题,提出一种输入增益补偿策略,并针对其通讯受限问题,提出一种带有死区操作器的事件触发控制机制.首先,采用伪偏导技术沿时间轴方向,在智能体的每个工作点上建立一种紧格式动态线性化数据模型,并给出该数据模型的参数估计算法.在此数据模型的基础上结合符号图论,研究智能体之间的合作与竞争关系,设计一种事件触发的数据驱动双向编队控制算法.最后,通过李雅普诺夫稳定性理论、矩阵理论以及压缩映射原理论证所提出算法的收敛性,并通过仿真实验和实物实验进一步验证该算法的正确性与有效性.As for the random delay issue of multi-input multi-output(MIMO)nonlinear discrete-time multi-agent systems without dynamics models,an input gain compensation scheme is proposed.For the limited communication problem,we propose an event-triggered mechanism with a dead-zone operator.Firstly,we establish a compact form dynamic linearization data model on each work point of the agent using the pseudo partial derivative technique and propose the corresponding parameter estimation approach.Based on the obtained data model,by combining with the symbolic graph theory and researching the cooperative-competitive relationships among agents,we propose an event-triggered data-driven bipartite formation control algorithm.Finally,we prove the convergence of the designed algorithm using the Lyapunov stability theory,matrix theory,and contracting mapping principle and further demonstrate the effectiveness and correctness of the developed algorithm through simulation studies and hardware tests.

关 键 词：多智能体系统 数据驱动控制 随机时延 事件触发 双向编队 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]