Adaptive consensus of two-time-scale multi-agent systems

Abstract

This paper studies the distributed consensus problem of a new class of multi-agent systems, where each agent is represented by the two-time-scale system. Two fully distributed, ϵ-dependent adaptive consensus protocols are designed, namely, the relative states based adaptive protocol and the relative outputs based adaptive protocol. By using the merit of the proposed ϵ-Lyapunov function, some sufficient consensus criteria in terms of well-conditioned linear matrix inequalities are derived. The upper bound of ϵ is also determined. A numerical example for the drying section of a paper converting machine with nine rolls verifies the effectiveness of the proposed results.