In this paper, we develop an approach for scheduling slabs at continuous casters in the steel industry. The scheduling approach incorporates specific constraints such as flexible production orders, material supply in batches and different setup types. We further introduce a continually adjustable casting width, which corresponds to a technological control parameter. We present a new MILP model formulation, which integrates slab design and scheduling. Solutions for the model are obtained by a greedy randomized adaptive search procedure. We analyze the applicability and performance of the approach in a numerical case study which is based on real world data. High-valued feasible production plans can be obtained in reasonable computing time. The approach is able to solve industry size problem instances in reasonable time. As compared to the status-quo, on average savings in the number of charges of 10.6 % are obtained.
展开▼
