Rescheduling Penalties for Economic Model Predictive Control and Closed-Loop Scheduling

The ability of online optimization to improve the dynamic performance of complex systems, particularly in production scheduling and process control applications, has led to a significant interest in, and adoption of, online optimization algorithms and techniques for a range of industries. However, these algorithms' ability to account for subtle features in complex systems may cause nonintuitive and rapid alterations to the optimal trajectory for the human operators often observing, evaluating, and, particularly in scheduling applications, executing the tasks prescribed by the optimization algorithm. In practice, these alterations are often restricted in some manner by the scheduling algorithm to alleviate operators' concerns and improve overall facility operation. Although numerous rescheduling penalties and constraints for production scheduling algorithms have been proposed, theoretical properties concerning the dynamics of these modified systems have been largely ignored. In this paper, we provide a gen...