Observer-based dynamic event-triggered consensus tracking for nonlinear multi-agent systems under switching topologies
摘要
This paper studies the consensus tracking problem of multi-agent systems (MASs) with nonlinear dynamics and intermittent communication under directed switching topologies. By utilizing the discrete outputs of each agent and its neighbors, we design unknown input observers (UIOs) integrated with an event-triggered mechanism (ETM) to estimate the relative states. Building on this, a class of observer-based dynamic event-triggered (DET) controllers is proposed. The triggering rules employ dynamic auxiliary variables to conserve communication resources and explicitly account for topology switching instants. Unlike existing works, a dual ETM is introduced, where observer and controller triggering functions operate independently. Rigorous analysis shows that consensus tracking can be achieved if the average dwell time (ADT) exceeds a computable bound, while Zeno behavior is excluded for both observers and controllers. To demonstrate the feasibility of the proposed theory, two numerical simulations are presented as the final part of this paper.