Consensus of Homogeneous Agents with General Linear Dynamics under Switching Communication Networks

TL;DR

This paper presents a new consensus control method for identical linear multi-agent systems with switching communication networks, ensuring synchronization under broad conditions with simpler criteria than existing methods.

Contribution

It introduces a structured gain matrix approach for linear agents that guarantees consensus in switching networks, including unstable systems, with less complex conditions.

Findings

Consensus achieved with sufficiently large gain matrix

Applicable to fixed, undirected, and directed switching graphs

Simpler consensus conditions compared to prior work

Abstract

This work addresses the synchronization/consensus problem of identical multi-agent system (MAS) where the agents' dynamics are linear and the communication network is arbitrarily switching among connected topologies. The approach uses a gain matrix of a special structure in a dynamic compensator for each agent. Under reasonable conditions, the approach ensures that consensus is reached when the gain is sufficiently large. This result holds for general linear systems including the case where agents have repeated unstable eigenvalues. The proposed controller structure can be seen as a special case of the existing MAS controller structures but offers consensus conditions that are simpler than existing results, especially for the case when the network is switching among connected graphs. The works shows its application to three communication settings: fixed graph, switching among undirected and connected graphs and switching among directed and connected graphs. An example is provided to illustrate the results.