Coordinated Non-Cooperative Distributed Model Predictive Control for Decoupled Systems Using Graphs * * Funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) - GRK 1856.

摘要

Abstract: This paper proposes a novel strategy for Non-Cooperative Distributed Model Predictive Control (Non-Coop. DMPC) of Networked Control Systems (NCS) consisting of dynamically decoupled agents where the coupling is achieved in the objective function or in the constraints. Moreover, the coupling is considered to be time-variant. We call this strategy Priority-Based Non-Coop. DMPC (PB-Non-Coop. DMPC). It is based on assigning priorities to the agents and the use of the coupling graph. Each PB-Non-Coop. DMPC is associated with a different agent and computes the local control inputs based only on its states and that of its neighbors. PB-Non-Coop. DMPC satisfies the prediction consistency property and reduces the overall computation time in comparison with existing Non-Coop. DMPC strategies in literature, thereby improves the overall behavior of NCS. We compare PB-Non-Coop. DMPC with centralized MPC as well as with another Non-Coop. DMPC strategy from literature.