Metaheurísticas para o problema de programação de tarefas em máquinas paralelas com tempos de preparação dependentes da sequência e de recursos

• Autor(a) principal: Kampke, Edmar Hell
• Data de Publicação: 2010
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: LOCUS Repositório Institucional da UFV
• Texto Completo: http://locus.ufv.br/handle/123456789/2603
• topic: Otimização combinatória; Heurísticas; Metaheurísticas; Programação de tarefas; Máquinas paralelas; Combinatorial optimization; Heuristics; Metaheuristics; Task scheduling; Parallel machines; CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO

The problems of scheduling jobs on parallel machines are important problems of combinatorial optimization. They involve routine problems in industries of small and large. This work addresses an unrelated parallel machine problem with machine and job sequence dependent setup times. The characteristic of this problem is that the amount of setup times do not only depend on the machine and job sequence, but also on a number of resources assigned, which can vary between a minimum and a maximum. Due to the combinatorial complexity of this problem, initially we propose an algorithm based on the GRASP metaheuristic, in which the basic parameter that defines the restrictiveness of the candidate list during the construction phase is self-adjusted according to the quality of the solutions previously found (reactive GRASP). Then, we propose an algorithm based on the Iterated Local Search (ILS) metaheuristic. This metaheuristic was proposed recently in the literature and is being applied successfully to several combinatorial optimization problems. In both algorithms, an intensification strategy based on the path relinking technique is used. This consists in exploring paths between elite solutions found by the algorithms. The results obtained by the proposed algorithms are compared with each other and with the best results available in literature. The analysis and discussion of results obtained with the algorithms, leads us to conclude that the metaheuristics show good performance, it is possible to improve, on average, 9,14% the results of the literature, which demonstrates the feasibility of using these algorithms to solve real problems existing in industries.