Avaliando algoritmos genéticos com diferentes estruturas populacionais em um problema integrado de dimensionamento de lotes e programação da produção

Abstract

Genetic Algorithm were evaluated using different population structures and genetic operators of recombination to solve the Integrated Lot Sizing and Scheduling Problem. Uni and multi-population structures with individuals hierarchically organized or not in trees, were applied in instances found in soft drink companies. In these industries the production process has two interdependent levels, one concerning tanks which storage raw material and other concerning production lines where the beverage is bottling. The lot sizing and scheduling of raw materials in tanks and beverages in lines have to be simultaneously determined. Almost all approaches presented better results than estimated cost by factory, standing out the multi-population approaches hierarchically structured in trees.