In this paper, the finite-time leader-following output consensus problem is investigated for a multi-agent system subjected to directed topology and disturbance. Firstly, a distributed finite-time extended state observer (DFTESO) is designed to provide estimate for unmeasured consensus errors and disturbances. The sufficient condition is provided in terms of linear matrix inequality (LMI) to ensure that the error dynamics of the DFTESO is convergent in finite time. Then, based on the designed observer, the leader-following consensus protocol based on dynamic gain method is proposed under the directed topology. Finally, an example on the multi-spacecraft system is provided to illustrate the effectiveness of the proposed methodology.

本文研究了有向拓扑和扰动的多智能体系统的有限时间跟随跟随输出共识问题。首先，设计了分布式有限时间扩展状态观察器（DFTESO），以提供对未测共识误差和干扰的估计。根据线性矩阵不等式（LMI）提供了充分的条件，以确保DFTESO的误差动态在有限的时间内收敛。然后，根据设计的观察者，提出了有向拓扑下基于动态增益方法的领导者遵循共识协议。最后，提供了有关多航天器系统的示例，以说明所提出方法的有效性。