Linking Control Strategy Design and Model Predictive Control

The purpose of this paper is to describe the importance of the underlying relationship between control strategy design and model predictive control. Successes and problems encountered when implementing model predictive control (MPC) on chemical processes have revealed that understanding this relationship provides insight into the nature of the process control problem. Model predictive control (MPC) has been used as an effective tool to gain the process control benefits that come from its ability to handle constraints, process interactions, and multiple time frames. The use of the MPC algorithm on a variety of chemical processes has led to insight on how to effectively use MPC along with traditional control strategy notions to improve process control. The development of control strategies using MPC has resulted in the typically reported benefits of increased throughput and reduced process variability. Several issues remain to be addressed. These include controller tuning, complex performance criteria, depth of integration of MPC with the regulatory control layer, redundant process information, and controller robustness to measurement loss or deterioration.