Data-Driven EMPC with Design of Terminal Ingredients

This paper introduces a novel data-driven economic model predictive control (EMPC) framework for linear time-invariant (LTI) systems, using Willems’ Fundamental Lemma to describe system behavior with persistently exciting data. Our method incorporates terminal ingredients, including a terminal cost and terminal constraint, to ensure recursive feasibility and asymptotic stability. We detail the design procedure for these terminal ingredients within the data-driven EMPC context. The paper concludes by confirming the strong duality property, feasibility and stability of the economically optimal equilibrium in the closed-loop system. The proposed EMPC framework is validated through its application to an isothermal stirred-tank reactor (CSTR), demonstrating its effectiveness.