書誌事項
- タイトル別名
-
- A data mining method for mechanical interpretation of topology optimization results
抄録
Structural optimization problems are often characterized as multiobjective problems. Among structural optimization, topology optimization has the most potential to explore high-performance designs, since it allows topological changes as well as changes in shapes during the optimization procedure. A multiobjective optimization problem usually offers a set of optimal solutions, called Pareto-optimal solutions, depending on the conflict of the objectives. In multiobjective topology optimization situations, decision makers face the challenging task of choosing a solution that best meets their needs, based on cumbersome comparisons and trial-and-error attempts among a set of Pareto-optimal solutions. One solution to solve this issue is the integration of data mining techniques with multiobjective optimization methods which can help decision makers get useful knowledge. In this paper, we propose a data mining technique in multiobjective topology optimization in which clustering and association rule analysis are sequentially applied to a Pareto-optimal solution set. First, clustering is applied in the design space, visualizing the results in the objective space. Detailed features in each cluster are then analyzed based on the concept of association rule analysis, so that characteristic substructures can be extracted from each cluster. Several numerical examples are provided to demonstrate the applicability of the proposed method.
収録刊行物
-
- 設計工学・システム部門講演会講演論文集
-
設計工学・システム部門講演会講演論文集 2018.28 (0), 2307-, 2018
一般社団法人 日本機械学会
- Tweet
キーワード
詳細情報 詳細情報について
-
- CRID
- 1390001288139309184
-
- NII論文ID
- 130007654270
-
- ISSN
- 24243078
-
- 本文言語コード
- ja
-
- データソース種別
-
- JaLC
- Crossref
- CiNii Articles
-
- 抄録ライセンスフラグ
- 使用不可