Um estudo sobre limites duais para o problema integrado de dimensionamento de lotes e sequenciamento da produção

• Ano de defesa: 2015
• Autor(a) principal: Silva, Tiago Tiburcio da [UNESP]
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Estadual Paulista (Unesp)
• Palavras-chave em Português: Matemática; Otimização matematica; Metodos de relaxação (Matematica); Problema de dimensionamento de lotes; Algoritmos; Mathematical optimization
• Link de acesso: http://hdl.handle.net/11449/136656; http://www.athena.biblioteca.unesp.br/exlibris/bd/cathedra/24-03-2016/000859866.pdf

Mathematics is present in our daily routine to tell time, count money, predict the weather. Many manufacturing companies deal with daily decisions in the manufacturing sector, lot-sizing and sequencing their production. However, the most usual is to take these decisions considering two independent problems, and not simultaneously, as it adds better results. In this work we integrate these decisions through a mathematical model that adds to the problem of lot sizing, sequencing decisions using constraints of the type MTZ and MCF. We also study these two formulations, considering the set up decisions explicitly and implicitly resulting in four di erent mathematical formulations for the integrated problem. We conclude that the MCF formulation with the explicit set up variable is stronger than the other formulations studied and the solutions of the instances of formulations based on constraints of MTZ type are strongly in uenced by the cutting planes and pre-processing included in the solver CPLEX. We aimed to derive primal and dual bounds for the integrated problem of lot sizing and sequencing of production. To obtain the primal bound we proposed a greedy heuristic. The dual bounds were obtained studying the Lagrangean and the Lagrangean / Surrogate relaxation and the methods used to solve the dual associates were the subgradient algorithm and Volume algorithm. The method with better performance was the dual Lagrangian / Surrogate solved by subgradient Algorithm for formulation with constraints MTZ type and explicit set up variable