Robust periodic economic MPC for linear systems

Economic Model Predictive Control differs from conventional tracking model predictive control by directly addressing a plants economic cost as the stage cost, consequently leading to better economic performance. This paper extends current economic model predictive control theory to linear time-invariant systems with periodic disturbances and cost functions, under mild assumptions. To ensure an increased region of attraction and to continuously guarantee feasibility of the controller despite changing economic conditions, a periodic terminal condition is used in place of terminal constraints. The approach draws on constraint tightening techniques in order to guarantee robust satisfaction of constraints as well as convergence of the controller. A Lyapunov based approach is used to show stability of the proposed controller and characterise a region about the optimal trajectory to which the system converges.