レイアウト設計のためのメタ戦略によるハイブリッド技法

概要

論文の詳細を見る
The problem with 2D facility layout planning can be descrided as follows. Each placement unit is a rectangular block having various shapes and the purpose is to place all the blocks without any overlap, and to meet the given requirements for flows or closeness ratings between blocks as well as possible. This paper deals with applications of the genetic algorithm (GA) and the simulated annealing algorithm (SA) to such a combinatorial optimization problem. We clarify the effectiveness of new techniques for applying the GA or SA and the effects of some parameters on GA or SA performance, based on numerical experiments. These results let us determine the parameter tuning of the GA or SA. Then, we propose a new hybrid method combining the advantage of the GA for global searches and the advantage of the SA for local searches. After tuning major parameters of the method sufficiently, we compal^e the performances of the three methods proposed here. It is shown that the hybrid method has the best performance.
一般社団法人日本機械学会の論文
1997-07-25