Aplicação de algoritmos genéticos no planejamento de embarque em um terminal de contêineres

• Ano de defesa: 2008
• Autor(a) principal: Piva, Marcio Luiz
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Uberlândia; BR; Programa de Pós-graduação em Engenharia Elétrica; Engenharias; UFU
• Palavras-chave em Português: Otimização; Terminal; Planejamento de embarque; Contêiner; Algoritmos genéticos; Otimização combinatória; Genetic algorithms; Optimization; Stowage planning; Container; CNPQ::ENGENHARIAS::ENGENHARIA ELETRICA
• Link de acesso: https://repositorio.ufu.br/handle/123456789/14378

In globalization times, the international trade becomes part of day-by-day of the people. With this continuous increasing of the importation and exportation activities, the logistic chain starts to have vital and indispensable role. Containers Terminal is an important component of the international logistic chain; multimodal interface that handles mainly with goods of the maritime modal (ship), become one of the responsible by the agility and the cost that the goods that are destined and came from foreign commerce reaches the final consumer. One of the more important tasks of the Container Terminal operational activities is the Stowage Planning, which can be characterized as a work of the optimize resources and schedules. The large amount of variables in these optimizations becomes decision support tools indispensable. Genetic Algorithms (GA) have being successfully used in complex problems, mainly which deterministic modeling is difficult, or if this approach brings computational times that becomes solutions infeasible. The proposal of this work was the implementation of the GA that could enclose the Container Terminal reality, with the complete and the large number of variables and restrictions known for the Stowage Planning process. In the GA developed was used also a routine that seems the local search process, called repaired routine, intending to speed up the convergence to the feasible solutions space, improving the general performance. With the container set used, the simulations results showed that the employed technique is appropriate, considering the characteristics of the best solutions and the computational times. Future works suggested would become the proposal still more adherent to daily run of the Container Terminal.