Zone Control DMC With Extremum Seeking: A Proposal for Improving Economic Goals

Abstract

This paper presents a novel approach to incorporate economic optimization into industrial control by integrating Zone Dynamic Matrix Control (DMC) with Extremum Seeking (ES) techniques. Zone DMC is employed to manage multivariable constrained processes, particularly for scenarios where maintaining variables within specified ranges is prioritized over precise setpoint tracking. In contrast, ES is utilized to optimize economic objectives without relying on an explicit process model, making it particularly valuable in complex or uncertain systems. The proposed integration leverages the strengths of both controllers, allowing ES to actuate only when all controlled variables are within predefined zones and ensuring seamless transitions between control modes. The effectiveness of the approach is demonstrated through numerical simulations, which illustrate improvements in economic performance even in the presence of measurement noise. Additionally, avenues for further research, including the development of stopping criteria for the ES controller and its integration with a Real-Time Optimization (RTO) layer, are discussed. Overall, this study showcases the potential of combining advanced control strategies to achieve superior economic outcomes in industrial processes.