Data-driven-based fully distributed event-triggered control for nonlinear multi-agent systems

Abstract

This paper investigates the fully distributed control issue for nonlinear multi-agent systems (MASs) under the limited network bandwidth. A novel event-triggered MFAC framework is developed to ensure the consensus of system output signals in a fully distributed manner, in which an adaptive step-size operator is introduced to eliminate the reliance on communication topology information. To save communication resources efficiently, a dynamic event-triggered mechanism (ETM) with switch-adjustable threshold parameters and dormancy waking functions is designed. This triggering mechanism not only accelerates the convergence of consensus errors in unstable situations, but also prolongs the inter-execution time while maintaining the desired control performance. In the entire control process, only measured input/output data are utilized. Through mathematical analysis, the average consensus errors of closed-loop MASs are proven to converge to zero asymptotically. Finally, the effectiveness and superiority of the proposed method are demonstrated through simulation comparisons.