Bipartite Control for Cooperative–Antagonistic Unknown Nonlinear Multiagent Systems With Link Faults

Abstract

This article aims at the problem of leader-following bipartite consensus (BI-consensus) control for heterogeneous multiagent systems with antagonistic interactions in the presence of communication link faults and unknown nonlinearities. First, a distributed adaptive communication policy is designed to compensate for the time-varying and unknown topological weights caused by communication faults, which solves the problem of strong coupling between communication faults and the Laplacian matrix. Second, the radial basis function neural networks are utilized to approximate the unknown nonlinear functions online to compensate for the system uncertainty. Furthermore, combining the neural network approximation mechanism and the adaptive communication policy of time-varying unknown weights, a novel fully distributed adaptive cooperative–antagonistic control strategy is presented to achieve the leader-following BI-consensus. The theoretical results show that the followers can reach the BI-consensus concerning the leader in both undirected and directed signed graphs. Two numerical examples are presented to verify the correctness and effectiveness of the scheme.