A Distributed Receding Horizon Control Algorithm for Dynamically Coupled Nonlinear Systems

Abstract

This paper presents a distributed receding horizon control law for dynamically coupled nonlinear systems that are subject to decoupled input constraints. The subsystem dynamics are modelled by continuous time nonlinear ordinary differential equations, and the coupling comes in the form of state dependence between subsystems. Examples of such systems include tight formations of aircraft and certain large scale process control systems. Given separable quadratic cost structure, a distributed controller is defined for each subsystem, and feasibility and asymptotic stabilization are established. Coupled subsystems communicate the previously computed trajectory at each receding horizon update. Key requirements for stability are that each distributed optimal control not deviate too far from the remainder of the previous optimal control, and that the amount of dynamic coupling is sufficiently small.