Modelos e algoritmos de otimização para agrupamento de pedidos e sequenciamento de panelas em lingotamento contínuo

Abstract

The melt shop is the sector responsible for the transformation of pig iron (from blast furnace) in steel and consists of three stages: primary refining, secondary refining and continuous casting (responsible for liquid steel solidification and its transformation into plates). The liquid steel is transported from one stage to the other by large capacityladles. One of the main problems encountered in the melt shop is the definition of orders that will compose a ladle and in the programming of these ladles in the caster, in order to minimize the costs of production of this company. However, it was observed, in literature researched, there are few papers that deal with the batching orders problem and sequencing charges (ladles) problem in an integrated manner. Most existing papers seeks to deal with the problems separately. In addition, few papers deal with the formation of intermix slabs versus perform a setup operation, one of the main problems found in most steel plants. So, in this thesis, it was proposed the use of mixed integer linear programming models and heuristics to solve efficiently the sequencing charges and batching orders problem in a steel plant, separately and in an integrated manner. The models and heuristics were implemented in language AMPL and solved with the optimization package CPLEX. To test the efficiency of the models and heuristics, it was created two sets of instances based on reports provided by a Brazilian steel plant. Theresults were satisfactory, enabling and encouraging the application of these methods in real-world scenario