Zhang Tian-Yu, Ye Dan, Zhao Xingang
IEEE Trans Cybern. 2022 Aug;52(8):7865-7874. doi: 10.1109/TCYB.2021.3052988. Epub 2022 Jul 19.
This article studies the leader-following consensus of multiagent systems (MASs) with completely event-triggered mechanisms (CETMs) and directed switching topologies. When most event-triggered schemes in MASs only reduce each agent's data transmissions to neighbor agents, CETMs further reduce data transmissions to actuators with one triggered function in each agent. To guarantee the switching links utilized by CETMs contain a directed spanning tree at any time, triggered decisions are improved to include the instants at which each agent's output links or leader link change. Furthermore, a less conservative method based on matrix inequalities is combined to minimize the allowable average dwell time (ADT) of switching topologies under the design conditions of CETMs. Based on multiple Lyapunov functions (MLFs), the sufficient conditions for controller gains that guarantee the leader-following consensus of MASs are proposed when the ADT of switching topologies is larger than the allowable value. Finally, an example of autonomous underwater vehicles (AUVs) is given to illustrate the effectiveness of CETMs.
本文研究了具有完全事件触发机制（CETMs）和有向切换拓扑的多智能体系统（MASs）的领导者-跟随者一致性问题。当多智能体系统中的大多数事件触发方案仅减少每个智能体向邻居智能体的数据传输时，完全事件触发机制通过在每个智能体中使用一个触发函数进一步减少向执行器的数据传输。为了确保完全事件触发机制所使用的切换链路在任何时候都包含一个有向生成树，对触发决策进行了改进，使其包括每个智能体的输出链路或领导者链路发生变化的时刻。此外，结合一种基于矩阵不等式的保守性较低的方法，在完全事件触发机制的设计条件下，最小化切换拓扑的允许平均驻留时间（ADT）。基于多重李雅普诺夫函数（MLFs），当切换拓扑的允许平均驻留时间大于允许值时，提出了保证多智能体系统领导者-跟随者一致性的控制器增益的充分条件。最后，给出了一个自主水下航行器（AUVs）的例子来说明完全事件触发机制的有效性。
