Estratégias Relax-and-Fix para o sequenciamento da produção de bobinas de aço

Abstract

The production scheduling problem in the steel industry is a complex challenge requiring decision-making under various operational and strategic constraints. These constraints often involve significant trade-offs that must be balanced to ensure both the efficiency and feasibility of the production process. This study proposes a deterministic approach to modeling and solving the production sequencing problem using Mixed-Integer Programming (MIP). The primary objective is to minimize costs related to order delays, early completions, and intermediate inventory during coil production, within a multi-stage production environment that incorporates precedence relationships between tasks. This work explores heuristic techniques based on MIP, such as relax-and-fix and strategies for partitioning large instances, aiming to reduce computational time and enable the application of MIP to larger-scale problems. The results demonstrate that, even in complex industrial environments such as the steel industry, it is possible to effectively apply integer programming models, particularly when combined with heuristic approaches that provide significant computational improvements.