Inferência fuzzy para o problema de corte de estoque com sobras aproveitáveis de material
Carregando...
Data
2011-04-01
Autores
Cherri, Adriana Cristina [UNESP]
Alem Junior, Douglas José
Silva, Ivan Nunes da
Título da Revista
ISSN da Revista
Título de Volume
Editor
Sociedade Brasileira de Pesquisa Operacional
Resumo
Nesse trabalho, e estudado o problema de corte de estoque com sobras aproveitaveis (PCESA), que consiste em determinar uma política de aproveitamento das sobras de objetos cortados. Estas sobras, desde que não sejam demasiadamente pequenas, podem retornar ao estoque como retalhos para atender demandas futuras, não sendo, portanto, consideradas perdas. Para resolver o PCESA, podem ser utilizados varios procedimentos heurísticos propostos na literatura. Entretanto, e difícil identificar o melhor método de solução, pois cada um deles apresenta uma solução diferente em relação as características de interesse do problema, como quantidade de retalhos produzidos, perdas e distribuição de perdas e retalhos nos objetos. Para superar essa dificuldade, propoe-se utilizar técnicas da lógica fuzzy para considerar a contributo de cada característica de interesse na solução, classifica-la de acordo com critérios pré-estabelecidos e, finalmente, identificar o melhor método com base na respectiva solução. Com a metodologia fuzzy desenvolvida, facilitou-se a análise dos métodos de solução
In this paper, we study the cutting stock problem with usable leftovers (CSPUL), that consists in determining a policy of usable leftovers from cut objects. These leftovers, if large enough, can returned to the stock as retails in order to fulfill future demands and, therefore, they are not computed as waste. In order to solve the CSPUL, several heuristic procedures from the literature can be used. However, it is hard to point out which solution method performs better, since each one presents different solutions regarding some conflicting characteristics, such as quantity of new retails generated, wastes, and distribution of the wastes and retails in the cut objects. To overcome this difficulty, we propose some fuzzy techniques to consider the contribution of each characteristic in the solution, classify it according some criteria beforehand considered and, finally, identify the best solution method based on the respective solution. With the proposed fuzzy methodology, the solution methods analysis was simplified
In this paper, we study the cutting stock problem with usable leftovers (CSPUL), that consists in determining a policy of usable leftovers from cut objects. These leftovers, if large enough, can returned to the stock as retails in order to fulfill future demands and, therefore, they are not computed as waste. In order to solve the CSPUL, several heuristic procedures from the literature can be used. However, it is hard to point out which solution method performs better, since each one presents different solutions regarding some conflicting characteristics, such as quantity of new retails generated, wastes, and distribution of the wastes and retails in the cut objects. To overcome this difficulty, we propose some fuzzy techniques to consider the contribution of each characteristic in the solution, classify it according some criteria beforehand considered and, finally, identify the best solution method based on the respective solution. With the proposed fuzzy methodology, the solution methods analysis was simplified
Descrição
Palavras-chave
problemas de corte de estoque, sobras aproveitaveis, inferencia fuzzy, cutting stock problems, usable leftovers, fuzzy inference
Como citar
Pesquisa Operacional. Sociedade Brasileira de Pesquisa Operacional, v. 31, n. 1, p. 173-195, 2011.