Robust economic model predictive control with zone tracking

TL;DR

This paper introduces a robust economic model predictive control method with zone tracking for uncertain nonlinear systems, ensuring finite-step zone achievement and improved economic performance through a risk-based tuning approach.

Contribution

It proposes a novel EMPC design that guarantees zone tracking within finite steps and incorporates a risk factor for economic optimization under uncertainties.

Findings

Successfully achieves zone tracking within finite steps

Demonstrates improved economic performance with risk tuning

Validates approach on a nonlinear chemical process example

Abstract

This paper presents a robust economic model predictive control (EMPC) formulation with zone tracking for discrete-time uncertain nonlinear systems. The proposed design ensures that the zone tracking objective is achieved in finite steps and at the same time optimizes the economic performance. In the proposed design, instead of tracking the original target zone, a robust control invariant set within the target zone is determined and is used as the actual zone tracked in the proposed EMPC. This approach ensures that the zone tracking objective is achieved within finite steps and once the zone tracking objective is achieved (the system state enters the robust control invariant set), the system state does not come out of the target zone anymore. To optimize the economic performance within the zone in the presence of disturbances, we introduce the notion of risk factor in the controller design. An algorithm to determine the economic zone to be tracked is provided. The risk factor determines the conservativeness of the controller and provides a way to tune the EMPC for better economic performance. A nonlinear chemical example is presented to demonstrate the performance of the proposed formulation.