Coordinated-distributed MPC of nonlinear systems based on price-driven coordination

The main objective of this work is to design a coordinated distributed model predictive control (CDMPC) architecture for nonlinear systems using the price-driven coordination method. One fundamental assumption of this work is that a centralized model predictive control (MPC) scheme can be designed based on successive linearization of the nonlinear system that is able to asymptotically stabilize the closed-loop system at the origin and the coordination schemes strives to get to the very same performance via adding a coordinator level to the existing decentralized structure. In other words, the coordinator and the distributed MPCs exchange information and calculate the optimal future input trajectories iteratively. In order to deploy the price-driven coordination algorithm developed by Marcos [1], which is applicable to linear or linearized systems, the alkylation process of benzene used as a case study was successively linearized around the operating points. The simulation results demonstrate that the algorithm can be successfully applied to nonlinear systems using a successive linearization strategy.