In this study, the distributed output consensus control issue is investigated for a class of linear cluster multi-agent systems (CMASs) under the control strategy of the reset observer. We consider a communication network consisting of several clusters, each of which is directed and contains a leader. The interactions among agents include continuous-discrete hybrid communication. Specifically, an instantaneous connectivity only exists between the clusters at discrete moments, called the reset time sequence. At the reset time, an instantaneous fixed directed network is formed such that only the leaders will consider the available information of neighboring leaders to reset their own states. During non-reset intervals, only the intra-clusters are connected while the inter-clusters are equivalent to a disconnected network topology. Considering that in practice, the state information may be partially unavailable, only the relative output information is utilized to estimate the unavailable state and thus control protocols are developed with the help of the reset full-order and reduced-order observers, respectively. The stability of the closed-loop CMAS at both the reset time and non-reset intervals is studied based on Lyapunov analysis. The consensus value depends only on the initial conditions and the network topology involved, and not on the reset time sequence. Finally, numerical simulations are provided to illustrate the theoretical results.

在这项研究中，研究了一类线性集群多智能体系统（CMAS）在重置观察者控制策略下的分布式输出一致性控制问题。我们考虑一个由几个集群组成的通信网络，每个集群都是有向的并包含一个领导者。代理之间的交互包括连续-离散混合通信。具体来说，瞬时连接仅存在于离散时刻的集群之间，称为重置时间序列。在重置时，会形成一个瞬时固定有向网络，只有领导者才会考虑相邻领导者的可用信息来重置自己的状态。在非重置间隔期间，只有集群内连接，而集群间相当于断开的网络拓扑。考虑到在实践中状态信息可能部分不可用，仅利用相对输出信息来估计不可用状态，从而分别在复位全阶和降阶观测器的帮助下开发控制协议。基于Lyapunov分析研究了闭环CMAS在复位时间和非复位间隔的稳定性。共识值仅取决于初始条件和所涉及的网络拓扑，与重置时序无关。最后，提供了数值模拟来说明理论结果。仅使用相对输出信息来估计不可用状态，因此分别在重置全阶和降阶观察器的帮助下开发控制协议。基于Lyapunov分析研究了闭环CMAS在复位时间和非复位间隔的稳定性。共识值仅取决于初始条件和所涉及的网络拓扑，与重置时序无关。最后，提供了数值模拟来说明理论结果。仅使用相对输出信息来估计不可用状态，因此分别在重置全阶和降阶观察器的帮助下开发控制协议。基于Lyapunov分析研究了闭环CMAS在复位时间和非复位间隔的稳定性。共识值仅取决于初始条件和所涉及的网络拓扑，与重置时序无关。最后，提供了数值模拟来说明理论结果。