Consensus in multi-agent networks with switching topology and nonlinear dynamics

Abstract

This paper investigates a double-integrator consensus problem for multi-agent networks (MANs) with nonlinear dynamics, where each agent is governed by both time-varying position and velocity consensus terms. The following three cases are carefully analyzed: 1) networks with fixed topology; 2) networks with controlled switching topology; and 3) networks with random switching topology. Based on previous work [11], some sententious sufficient criteria are established for reaching consensus in MANs with fixed topology. With the help of consensus results under fixed topology, some criteria are also derived to achieve consensus for MANs with switching topology by utilizing average dwell time approach and a programmable switching signal. The next moment, in view of ergodic property and stationary distribution of Markov chain, some criteria are derived to reach consensus for MANs with random switching topology. It is worth noting that the MANs considered in the above two switching cases are composed of all or partial individual switching topologies which can be reached the desired consensus.