Multi-agent consensus with heterogeneous time-varying input and communication delays in digraphs
Wei Jiang,Kun Liu,Themistoklis Charalambous,Wei Jiang,Kun Liu,Themistoklis Charalambous
DOI: https://doi.org/10.1016/j.automatica.2021.109950
IF: 6.4
2022-01-01
Automatica
Abstract:This paper investigates the distributed consensus tracking control problem for general linear multi-agent systems with external disturbances and heterogeneous time-varying input and communication delays under a directed communication graph topology, containing a spanning tree. First, for all agents whose state matrix has no eigenvalues with positive real parts, a communication-delay-related observer, which is used to construct the controller, is designed for followers to estimate the leader’s state information. Second, by means of the output regulation theory, the results are relaxed to the case that only the leader’s state matrix needs to be asymptotically stable or marginally stable and, under these relaxed conditions, the controller is redesigned. Both cases lead to a closed-loop error system of which the stability is guaranteed via a Lyapunov–Krasovskii functional with sufficient conditions in terms of input-delay-dependent linear matrix inequalities (LMIs). An extended LMI is proposed which, in conjunction with the rest of LMIs, results in a solution with a larger upper bound on delays than what would be feasible without it. It is highlighted that the integration of communication-delay-related observer and input-delay-related LMI to construct a fully distributed controller (which requires no global information) is scalable to arbitrarily large networks. The efficacy of the proposed scheme is demonstrated via an illustrative numerical example.
automation & control systems,engineering, electrical & electronic