Production Scheduling of a Large-Scale Industrial Batch Plant. II. Reactive Scheduling

Abstract

The aim of reactive scheduling is to update the current production schedule in order to provide an immediate response to an unexpected event such as equipment breakdown or the addition/modification of orders. Reactive scheduling formulations should be fast and efficient and take into account the schedule currently in progress as well as planned productions that are not affected by the unexpected event. In this work, a reactive scheduling framework is developed which utilizes an efficient mixed-integer linear programming (MILP) mathematical framework developed for short-term scheduling problems with modifications introduced to reflect the effects of the unforeseen event. To avoid full rescheduling of the current production horizon, the formulation determines tasks which are not affected by the unforeseen event, either directly or indirectly, and can be carried out as scheduled. The resulting tasks along with additional subsets of tasks are then fixed in the MILP problem, and the rest of the horizon is rescheduled. We consider two types of unexpected events: unit shutdown and the addition or modification of orders. Both cases utilize the nominal schedules, the original MILP formulation for short-term scheduling with modifications, and a program to determine which tasks may be fixed before rescheduling. The formulation is then able to determine an updated production schedule for the remaining time horizon in a reasonable amount of CPU time. Reactive scheduling of a large-scale industrial batch plant is performed to demonstrate the effectiveness of the proposed approach.