Two-Level Hierarchical Mission-Based Model Predictive Control

A two-level hierarchical model predictive control (MPC) formulation is presented for constrained linear systems operating over a mission. Mission-based MPC is applicable to many control applications where the system operates for a finite time and stability about an equilibrium is not the primary objective. Instead, the primary control objective is to guarantee constraint satisfaction during operation as well as terminal constraints imposed on the final state of the system at the end of the mission. The secondary control objective is reference tracking, where references determine the desired operation for the system. A hierarchical control formulation permits the upper level controller to plan state trajectories over the entire mission, while a lower level controller modifies these trajectories to improve reference tracking. This decomposition of the control problem reduces computational cost, enabling real-time implementation for large systems with long missions. Feasibility proofs guarantee the constraint satisfaction while a numerical example demonstrates the efficacy of the approach.