1779. Integrated Scheduling of Continuous Casting and Hot Rolling in Steel Production: Modeling as a Two-stage Lot Sizing/Assignment Problem
Invited abstract in session MB-49: Integrated lot-sizing problems, stream Lot Sizing, Lot Scheduling and Production Planning.
Monday, 10:30-12:00Room: M1 (building: 101)
Authors (first author is the speaker)
| 1. | Hubert Missbauer
|
| Information Systems, Production and Logistics Management, University of Innsbruck | |
| 2. | Alexander Lohr
|
| Information Systems, Production and Logistics Management, University of Innsbruck | |
| 3. | Julian Fodor
|
| Information Systems, Production and Logistics Management, University of Innsbruck |
Abstract
Scheduling of the sequential production stages Continuous Casting and Hot Rolling in steel production for a time horizon of about 7 days is a complex problem that requires considering technical constraints at both stages as well as limited capacities of upstream facilities (steelmaking and refining) and intermediate stages (slab processing and heat preservation pits). Due to this complexity, decomposition is required to handle the problem. Based on the case of an Austrian steel plant we elaborate on the decomposition principles applicable, mainly temporal decomposition and two-stage block planning, where the former seems to be closer to practical applicability. In our research we pursue the latter. We show that the technical characteristics of the two stages lead to a combined two-stage lot sizing and slab assignment problem that can be solved by a fix-and-relax heuristic using an MILP solver. The sequencing problem on the continuous casters and the determination of start and completion times of the casts and of the rolling turns (timing) must be solved subsequently due to the problem decomposition. Numerical experiments indicate that the sequencing problem can be solved well by established methods provided that appropriate constraints are formulated at the lot sizing/assignment stage. The timing problem can be solved by resource-constrained project scheduling methods, but the capacity constraint of slab processing requires extension of the scheduling method we selected.
Keywords
- Scheduling
- Optimization Modeling
- Programming, Mixed-Integer
Status: accepted
Back to the list of papers