Modelos e algoritmos para problemas de localização em logística reversa

• Ano de defesa: 2017
• Autor(a) principal: Ellen Kenia Fraga Coelho
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; Brasil; ENG - DEPARTAMENTO DE ENGENHARIA PRODUÇÃO; Programa de Pós-Graduação em Engenharia de Produção; UFMG
• Palavras-chave em Português: Logística reversa; Testes de fixação de variáveis; Decomposição de Benders; Algoritmos; Engenharia de produção
• Link de acesso: http://hdl.handle.net/1843/41773

Product remanufacturing is one of the most pro table activities in reverse logistics. Running a business plan, in which companies take responsibility for the waste generated at their end-of-life products, involves important strategic decisions. One of the challenges in planning the reverse ow of products is deciding where the reprocessing facilities should be installed. This decision directly in uences the transport variable costs and the xed costs. This work proposes two di erent models for reverse logistics location problems. The rst one is the Capacitated Plant Location Problem in Reverse Logistics (CPL-RL), in which we assume that the o ered material of each collection center is taken to a single facility for reprocessing. This assumption includes speci c cases where there is no logistic availability in the network to send the collected material to di erent locations. Mixed Integer Linear Problem (MILP) is solved by using a two steps algorithm. In the rst step, reduction tests are performed to determine a priori which facilities will be opened /closed. If all facilities are xed opened or closed then the solution is optimal. If not all facilities can have their status de ned in this way, the resultant problem has a fewer variables and it is solved using a Benders method. The dataset was randomly generated and the results showed that the applied techniques are appropriate, achieving the optimal solution for all test problems. The second problem is a Closed Looping Supply Chain model (CLSC). It combines characteristics of some classic models of literature and speci c legislation.