Consensus control for a network of high order continuous-time agents with communication delays

This paper is concerned with the consensus control problem for multi-agent systems with agents characterized by high-order linear continuous-time systems subject to communication delays between neighbouring nodes in the network. A new consensus protocol is proposed. It requires communication between neighbouring agents only at certain sampling points, rather than at all times. It is also unique in the sense that it is nonlinear in the continuous-time domain but linear when the agents are viewed in the sampled-data domain. Under the proposed consensus protocol, marginally stable multi-agent systems can reach consensus for any large delay. Unstable multi-agent systems achieve consensus when the time delay is within a certain range. Moreover, in the single-input case, we give an optimal control gain which yields the fastest consensus speed. The proposed technique is expected to pave a new way for new theoretical studies on network properties required for consensus control. (C) 2017 Elsevier Ltd. All rights reserved.