Robust $H_{\infty}$ Receding Horizon Control for a Class of Coordinated Control Problems Involving Dynamically Decoupled Subsystems

Abstract

This paper presents a robust receding-horizon-control (RHC)-based scheme for a class of coordinated control problems involving dynamically decoupled subsystems which are required to reach a consensus condition in some optimal way. A general case of constrained subsystems having possibly uncertain time-varying dynamics in addition to external disturbances is considered and a suitable H∞-based near-consensus condition is defined as the target condition to be achieved. The proposed scheme employs computationally efficient subsystem-level RHC policies in the H∞-based minmax-cost framework together with a distributed subgradient-based method to optimize and update the consensus signal and the subsystem control inputs at regular intervals. Furthermore, the proposed framework allows the incorporation of computational delays in the RHC policy formulation so that the desired control performance is always guaranteed. The performance of the proposed control scheme is illustrated with some simulation examples.