基于事件触发的不确定多智能体系统自适应一致性

Event-triggered adaptive consensus control for uncertain multi-agent systems

中文关键词  一般非线性多智能体系统  事件触发控制  模型参考自适应  参数不确定

英文关键词  general nonlinear multi-agent systems  event triggered control  model reference adaptive  uncertainties

中文摘要

      针对有向拓扑下一般非线性多智能体系统中参数不确定和系统通信能力受限等问题,本文提出了一种结合模型参考自适应和事件触发机制的一致性控制协议.该协议首先为每个参数不确定的智能体设计一个参考智能体,其中自适应律的设计保证了不确定参数的有效估计,然后基于事件触发机制向邻居智能体发送参考智能体状态,最后基于接收的邻居参考智能体状态及自身实际状态设计控制器.采用对参考智能体设计事件触发策略的方式,可以有效避免智能体间的连续通信,减少系统通信资源.利用矩阵论、代数图论以及Lyapunov稳定性理论,证明在该控制协议下,含参数不确定一般非线性多智能体系统能实现一致性,且不存在Zeno行为.仿真实例进一步验证了理论结果的有效性.

英文摘要

      For the problems of uncertain parameters and limited system communication ability in general nonlinear multi-agent systems under directed graphs, a consensus control protocol combining model reference adaptation and eventtriggered mechanism is proposed. The protocol first designs a reference agent for each uncertain agent, the design of adaptive law ensures the effective estimation of uncertain parameters, then sends the reference agent state to neighbor agents based on an event-triggered mechanism, and finally designs a controller based on the received neighbor reference agent state and its actual state. This protocol adopts the method of designing event trigger strategy based on the reference model of the agent to effectively avoid continuous communication between agents and reduce system communication resources. The matrix theory, algebraic graph theory, and Lyapunov stability theory are used to prove the control strategy. General nonlinear uncertain multi-agent systems can achieve consensus without Zeno behavior. Simulation examples further verify the validity of the theoretical results.