Bipartite Consensus Control for Multi-Agent Systems With Intermittent Communication: A Data-Driven Method

Abstract

This study explores the data-driven control methods to address the bipartite consensus of continuous-time multi-agent systems (MASs) under intermittent communication (IC). A key advantage of this approach is that it removes the requirement for prior knowledge of system parameters. Since MASs operate in an IC environment, their states may diverge during the non-communication intervals. To overcome this challenge, a more flexible relationship between the lengths of communication and non-communication intervals and the convergence rate is established. Remarkably, this approach eliminates the restrictions on communication and non-communication intervals in existing researches and offers a more adaptable description of IC. Both bipartite consensus conditions and control gains are obtained in terms of input-state data. Finally, an example is presented to illustrate the effectiveness of the proposed data-driven approach.