Robust Consensus Design of Uncertain Multiagent Systems With Bounded Gains and Incremental Nonlinear Interactions
Sabato Manfredi · IEEE Transactions on Industrial Informatics · 2024
This article presents design conditions to address the robust leader–follower consensus problem in uncertain multiagent systems (MASs) with bounded gains and nonlinear interactions. The gains are restricted to an interval of values determined by either parameter uncertainties or the constraints of the agents, inherent in an industrial setting. In addition, we model the unavoidable nonlinear uncertainties in MAS and its communication architecture using incremental sector-bounded time-varying nonlinear functions, which are particularly well-suited for relative state information-based cooperative protocols. By leveraging Lyapunov stability theory and the linear matrix inequality (LMI) technique, we establish LMI conditions for the distributed design of control gains for both leader and followers. The proposed LMIs exhibit scalability, as they remain independent of the number of nodes and do not necessitate global information regarding the Laplacian eigenvalues, making them well-suited for the design of industrial large-scale MASs. Finally, we numerically and experimentally validate the proposed design approach for paradigmatic scenarios: industrial warehouse inventory management systems and leader–follower algorithm for smart grid and decision-making management applications.