データ包絡分析法による設計改善案構築方法の提案 : チャイルドシートへの適用事例(機械力学,計測,自動制御)

概要

論文の詳細を見る
One of the important aspects in preliminary design is to make their products perform well without any imperfections when it spread out in the market. Therefore, we need to evaluate the products with the same kinds of products from many aspects. These evaluations can regard as multi- objective design optimization, and most likely scalar methods are used, such as the weighted sum method. In these methods, it is quite important to give optimal weight for each objective function, so that it fits with preference of the designers. Moreover, these optimal weights give us an important information of the improvements of design quantitatively. In this paper, we use data envelopment analysis(DEA), which can derive total evaluation and optimal weights from given datum from the products in the same kinds. Hence, we can derive the amount of improvement needed compared with the other products. To examine the effectiveness of the proposed method, we collect data of Ministry of Land, Infrastructure and Transport Government of Japan in 2002 and evaluate many child-seats. From the results, we could clarify a part of child seat, which should be improved, and showed its effectiveness through its considerations.
一般社団法人日本機械学会の論文
2005-06-25