Distributed Event-Triggered Consensus Control from Noisy Data Using Matrix Polytopes

This paper presents a novel data-driven polytopic approach to event-triggered consensus control of unknown leader-following multi-agent systems (MASs). A distributed data-driven event-triggered consensus control protocol is proposed that utilizes noisy input-state data to enable all followers to track the leader while reducing communication and computational burden. Unlike previous research that relies on quadratic matrix inequalities to characterize system uncertainties, this paper devises a data-based polytopic representation for MASs, which enables addressing the consensus control problem without using explicit system matrices. Based on this representation, a data-based criterion is established, utilizing matrix polytopes to ensure the asymptotic stability of the closed-loop MAS. Moreover, a co-design method is presented for the distributed controller gain and the triggering matrix, using only data and expressed in terms of linear matrix inequalities. Finally, numerical simulations are conducted to demonstrate the validity and effectiveness of the proposed data-driven approach.