FURTHER IMPROVEMENT OF FINITE-TIME CONSENSUS PROTOCOLS FOR DETAIL-BALANCED NETWORKS

Abstract

This paper presents a new class of protocols to solve finite-time consensus for multi-agent systems. The protocols are induced from the classical finite-time consensus algorithm by using the so-called protocol function. Sufficient conditions are established for networked agents to experience finite-time consensus under time-varying undirected and fixed directed topologies. Numerical simulations show that the proposed protocols can provide more flexibility to improve convergence rate.