A genetic algorithm-based heuristic with unequal individuals to tackle the lot-sizing and scheduling problem

• Tipo: Artigo em Livro de Atas de Conferência Internacional
• Data: 2013

Autores

Marcos Furlan (Autor) FEUP; Bernardo Almada Lobo (Autor) FEUP; Maristela Santos (Autor) FEUP; Reinaldo Morabito (Autor) Outra

Ata de Conferência Internacional

Título: Proceedings of XLV SBPO - Simpósio Brasileiro de Pesquisa Operacional

Páginas: 1-12

XLV SBPO - Simpósio Brasileiro de Pesquisa Operacional

Natal Brasil, 16 a 19 de setembro de 2013

Classificação Científica

FOS: Ciências da engenharia e tecnologias

CORDIS: Ciências Tecnológicas

Outras Informações

Resumo (PT): In this paper we present a novel genetic algorithm-based approach to tackle a lot-sizing and scheduling problem. Mathematical models, commonly used to represent these problems, have the planning horizon divided in periods and sub-periods. Adjusting the right number of sub-periods becomes a hard task, since higher values create flexible environments and smaller values reduce the problem size and problem complexity. The key idea of our method is to take some advantages about adjusting the number of sub-periods. In this way, a genetic algorithm which allows for the use of individuals with different size was developed. This feature was performed to maintain the flexibility, as well as, reducing the computational load. Preliminary results show the benefits obtained when variable number of sub-periods is used, nevertheless, the tightness of the lower bounds becomes more promising when the number of products increases. More studies are necessary to determine an ideal lower bound rule.

Idioma: Inglês

Tipo (Avaliação Docente): Científica