Cooperative Distributed MPC using the Alternating Direction Multiplier Method

摘要

We consider cooperative distributed model predictive control of a linear, time-invariant, discrete-time plant, which consists of coupled subsystems. The cooperating controllers minimize a quadratic cost criterion subject to input and state constraints. We outline two approaches for such distributed controllers using the alternating direction multiplier method, which allows to obtain convergence guarantees. We outline suitable stability criterions and illustrate the applicability and performance of the proposed methods by a simple example.