Consensus Tracking of Data-Sampled Nonlinear Multi-Agent Systems With Packet Loss and Communication Delay

Abstract

This paper investigates the tracking consistency regulation of high-order multi-agent systems (MASs) subject to Lipschitz nonlinear perturbations. Rooting at the leader node, the interagent interaction is described by a directed graph that contains a directed spanning tree. On designing the tracking protocol, an aperiodic data sampling scheme is introduced such that in-neighbors' information can be aperiodically sampled and transmitted via broadcast channels, where packet loss and communication delay are unavoidably existent. Thanks to algebraic graph theory and Lyapunov-Krasovskii functional technique, the concerned consensus tracking issue is cast into asymptotically stabilizing a class of switched nonlinear systems which consist of delayed and non-delayed components. It is proved that, moreover, the asymptotic stabilization of such systems essentially depends on packet loss characteristics. By explicitly characterizing the interplay among sampling interval, communication delay and network structure, consensualization criteria are formulated within the linear matrix inequality (LMI) framework. The design method is validated by numerical simulations.