Uma Heurística Baseada em Programação Dinâmica para o Problema de Corte Bidimensional Guilhotinado 2-Estágios

ABSTRACT

Cutting and packing problems are part of the production planning process in many industries (e.g. paper, glass, furniture). In some furniture industries, large rectangular objects have to be cut into smaller rectangles and there is a limited storage space for work in process. In this case there is interest in solving the constrained two-dimensional two-stages guillotine cutting problem (PCBG-2est). Several authors applied dynamic programming algorithms for solving the unconstrained two-dimensional cutting problem. However, for the constrained case this technique still presents some challenges due to the size of the state space. We propose a heuristic based on the two-step method of Gilmore and Gomory for the constrained PCBG-2est considering special constraints associated with the cutting equipment. The results of a computational study with three sets of instances show the efficiency of the proposal. In particular, for instances that are similar to the furniture industry, solutions were obtained with an average maximum gap of 4.4%.

Keywords:
Two-dimensional two-stages guillotine cutting problem; constrained problem; dynamic programming; heuristic