This paper deals with a warehousing problem in three dimensional space, by taking account of safety, availability of the capacity, and easiness for taking boxes out. This problem is a complicated combinatorial optimization problem with three-dimensional constraints. The objective is to get the most favorite arrangement of boxes in a warahouse. An important constraint is that the area of the base of the upper box is always smaller than that of the lower box. In the problem formulation, this

more »

... raint is treated explicitly, but the other constraints are included in the objective as the respective penalty functions. The genetic algorithm (GA) is applied to solving the problem. For searching for a solution efficiently, several procedures such as the individual description, the decoding rule, crossover and mutation are proposed for this problem. Particularly, a new idea of "column" is introduced in this paper. The column is an array of boxes arranged in the vertical direction, which is applied to converting the genotype into the phenotype without getting an infeasible solution of the problem. Numerical results are shown to investigate the effectiveness of these procedures.