IEEE Trans Cybern. 2023 Jul;53(7):4446-4458. doi: 10.1109/TCYB.2022.3182036. Epub 2023 Jun 15.
This article addresses the scaled consensus problem for a class of heterogeneous multiagent systems (MASs) with a cascade-type two-layer structure. It is assumed that the information of the upper layer state components is intermittently exchangeable through a strongly connected communication network among the agents. A distributed hierarchical hybrid control framework is proposed, which consists of a lower layer controller and an upper layer one. The lower layer controller is a decentralized continuous feedback controller, which makes the lower layer state components converge to their target values. The upper layer controller is a distributed impulsive controller, which enforces a scaled consensus for the upper layer state components. It is proved that the two layer controllers can be designed separately. By considering the dwell-time condition of impulses and the feature of the strongly connected Laplacian matrix, a novel weighted discontinuous function is constructed for scaled consensus analysis. By using the Lyapunov function, a sufficient condition for scaled consensus of the MAS is derived in terms of linear matrix inequalities. As an application of the proposed distributed hybrid control strategy, a relaxed distributed hybrid secondary control algorithm for dc microgrid is obtained, by which the balance requirement on the communication digraph is removed, and an improved current sharing condition is obtained.
本文针对一类具有级联型两层结构的异构多智能体系统（MASs）的比例一致性问题进行了研究。假设通过各智能体之间的强连通通信网络，可以间歇性地交换上层状态分量的信息。提出了一种分布式分层混合控制框架，它由下层控制器和上层控制器组成。下层控制器是一个分散的连续反馈控制器，可以使下层状态分量收敛到目标值。上层控制器是一个分布式脉冲控制器，可以使上层状态分量达到比例一致性。证明了可以分别设计这两层控制器。通过考虑脉冲的停留时间条件和强连通拉普拉斯矩阵的特征，为比例一致性分析构建了一个新的加权不连续函数。通过使用 Lyapunov 函数，以线性矩阵不等式的形式推导出 MAS 达到比例一致性的充分条件。作为所提出的分布式混合控制策略的应用，获得了松弛的直流微电网分布式混合二次控制算法，该算法消除了对通信有向图的平衡要求，并获得了改进的电流共享条件。
