2018 Fiscal Year Final Research Report
Visualization and mathematical modeling of latent structure between global/local design information
Project/Area Number |
16K00295
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Multi-year Fund |
Section | 一般 |
Research Field |
Intelligent informatics
|
Research Institution | The University of Electro-Communications |
Principal Investigator |
CHIBA Kazuhisa 電気通信大学, 大学院情報理工学研究科, 准教授 (50450705)
|
Project Period (FY) |
2016-04-01 – 2019-03-31
|
Keywords | データ解析 / 進化計算 / 航空宇宙工学 / 数値流体解析 |
Outline of Final Research Achievements |
Beneficial information for designing engineering products is classified into global and local design information. Because global design information states the knowledge from past experiences clearly, physical interpretation is easy but there is not any innovations. Local design information can include hypotheses to innovative design, but its physical interpretation is difficult since machine learning algorithms indicate it. In this research, I have proposed a system to structurally visualize a latent structure between global and local design information. Previously, each information was dominantly extracted, but the approach has made it be available to exhaustively systematize. The results make physical interpretation of local design information easy and lead mathematical modelings.
|
Free Research Field |
設計情報学
|
Academic Significance and Societal Importance of the Research Achievements |
物理的解釈の困難だった局所設計情報に物理的解釈を加える一手法を創出したことにより,経験知として蓄積できなかった,設計要求を改善するための全く新しい設計情報を獲得でき,革新設計に至る仮説を導くことができるようになった.また,これらの情報を実問題設計にフィードバックすることで,より利便性の高い新たな製品の創出につながり,社会貢献につながるものと期待される.少なくとも本研究内で適用した次期宇宙輸送機の設計では,50年以上停滞しているコンセプトの刷新につながるとともに,今後航空産業への発展にも影響を与えることに鑑みれば,波及性の強い研究成果である.
|