Leader-following consensus of multi-agent systems under antagonistic networks

Abstract

In this paper, we address the leader-following consensus of multi-agent systems (MASs) under a class of antagonistic networks. Different from the traditional literature, in our model, whether a follower can access the leader depends on a matrix instead of a scalar. For antagonistic and directed networks with structurally balanced properties, sufficient conditions are established under which the MASs will achieve bipartite consensus. Detailed algorithms are proposed to design the gain matrices and the coupling strength. What is more, we design adaptive laws to self-tune the value of the coupling strength. Simulation results are also given to verify the effectiveness of our methods.