Otimização por algoritmos genéticos do sequenciamento de ordens de produção em ambientes Job Shop

• Ano de defesa: 2014
• Autor(a) principal: Grassi, Flávio
• Orientador(a): Pereira, Fabio Henrique
• Banca de defesa: Morabito Neto, Reinaldo , Araújo, Sidnei Alves de
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Nove de Julho
• Programa de Pós-Graduação: Programa de Pós-Graduação de Mestrado e Doutorado em Engenharia de Produção
• Departamento: Engenharia
• País: BR
• Palavras-chave em Português: job shop; algoritmo genético; otimização; sequenciamento da produção
• Palavras-chave em Inglês: job shop; genetic algorithms; optimization; production scheduling
• Área do conhecimento CNPq: ENGENHARIAS::ENGENHARIA DE PRODUCAO
• Link de acesso: http://bibliotecatede.uninove.br/tede/handle/tede/212

The optimization of processes is a highly relevant topic in the industry, and therefore treated by many researchers around the world for more than fifty years. Among the problems to be solved, must highlight the issue of sequencing of the production scheduling due its wide applicability, such as increasing productivity of vehicles in automobile industry or improving the performance of processors in computers. In the present work were conducted studies involving scheduling of production orders of deterministic problems in Job Shop environments through the use of genetic algorithms. The problems tested belong to a group available from an operations research library, widely used by researchers in this context, and in such instances all production orders are available for allocation since the instant zero, and processing times are fixed. The fitness function of the solutions generated during optimization was developed in C language. The adopted genetic algorithm uses conventional genetic operators and binary representation, and promotes improvements in relation to research which are also based on these operators, through the choice of the initial population, which is performed by adopting a concept of dynamic seed developed in this dissertation. Initially the generating seed of the population in the genetic algorithm adopts a simple sequencing rule based on the sequence of the production orders in relation to its route, which is defined by the problem, and then new seeds are used, which are those that generated the best individuals of previous generations. As presented in the results section, this concept of dynamic seed effectively generates a larger number of feasible solutions. The qualitative results show that the developed approach is competitive in relation to other classic representations of Job Shop environments, providing solutions in acceptable time for this sort of problem.