Modelos de programação matemática para problemas de carregamento de caixas dentro de contêineres

• Autor(a) principal: Junqueira, Leonardo
• Data de Publicação: 2009
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da UFSCAR
• Texto Completo: https://repositorio.ufscar.br/handle/ufscar/3600
• topic: Pesquisa operacional; Sistema de unitização de cargas; Otimização combinatória; Modelagem matemática; Carregamento de contêineres; Problemas de corte e empacotamento; Cutting and packing problems; Container loading; Combinatorial optimization; Mathematical modeling; ENGENHARIAS::ENGENHARIA DE PRODUCAO

The object of this study is a particular case of the cutting and packing problems, known as container loading problems. These problems consist in arranging rectangular boxes orthogonally into containers (or into trucks, railcars and pallets), in order to optimize an objective function, for example, maximize the utilization of the available space, or minimize the number of the required containers to load all the available items. The objective of this study is to develop mathematical programming models to deal with situations commonly found in container loading practice. Multiple orientations of the boxes, weight limit of the container, cargo stability, load bearing strength of the boxes and multiple destinations of the cargo are considered. The author is not aware of mathematical formulations available in the cutting and packing literature that deal with such considerations, and this paper intends to contribute with possible formulations that describe these situations, although not very realistic for being used in practice. Computational experiments with the proposed models are performed with the software AMS/CPLEX and randomly generated instances extracted from the cutting and packing literature. The results show that the models are consistent and properly represent the practical situations treated, although this approach (in its current version) is limited to solve to optimality only medium-sized problems. However, we believe that the proposed models can be useful to motivate future research exploring decomposition methods, relaxations, heuristics, among others, to solve the present problems.