Robust Decentralized MPC Algorithm for a class of Dynamically Interconnected Constrained Systems

Model predictive control (MPC) is one of the most promising and significant control approaches because of its abillity to handle control problems for constrained systems. In this paper, we propose a robust decentralized MPC algorithm for dynamically inter connected large-scale constrained uncertain systems with disturbances. Each decentralized MPC controller determines the controls equence, satisfying in put and state constraints in the presence of uncertainties and disturbances, based only on the nominal model of its own subsystem and information broadcasted from other controllers. This controller uses the information on the reachable state bound of its own subsystem and dynamically interconnected subsystems which is predicted one-step before. This algorithm guarantees robustst ability in the presence of model uncertainties and disturbances.