Application of integer programming on logistics solution for load transportation: the solver tool and its limitations in the search for the optimal solution

• Autor(a) principal: Santos, Ricardo França
• Data de Publicação: 2012
• Outros Autores: Souza Junior, Eugênio Correa, Bouzada, Marco Aurélio Carino
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Revista Produção Online
• Texto Completo: https://www.producaoonline.org.br/rpo/article/view/866
• topic: Pesquisa Operacional; Programação Inteira; Solver; Logística de Distribuição; Otimização de Carga.

This work tries to solve a typical logistics problem of Navy of Brazil regards the allocation, transportation and distribution of genera refrigerated for Military Organizations within Grande Rio (RJ). After a brief review of literature on Linear/Integer Programming and some of their applications, we proposed the use of Integer Programming, using the Excel’s Solver as a tool for obtaining the optimal load configuration for the fleet, obtaining the lower distribution costs in order to meet the demand schedule. The assumptions were met in a first attempt with a single spreadsheet, but it could not find a convergent solution, without degeneration problems and with a reasonable solution time. A second solution was proposed separating the problem into three phases, which allowed us to highlight the potential and limitations of the Solver tool. This study showed the importance of formulating a realistic model and of a detailed critical analysis, which could be seen through the lack of convergence of the first solution and the success achieved by the second one.