| Project/Area Number |
16K06088
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Fluid engineering
|
|---|
| Research Institution | Nagasaki University |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2016-04-01 – 2019-03-31
|
| Project Status |
Completed (Fiscal Year 2018)
|
| Budget Amount *help |
¥4,810,000 (Direct Cost: ¥3,700,000、Indirect Cost: ¥1,110,000)
Fiscal Year 2018: ¥1,690,000 (Direct Cost: ¥1,300,000、Indirect Cost: ¥390,000)
Fiscal Year 2017: ¥2,080,000 (Direct Cost: ¥1,600,000、Indirect Cost: ¥480,000)
Fiscal Year 2016: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
|
|---|
| Keywords | 流体機械 / 人工知能 / 最適化 / サージ / ケーシングトリートメント / 小弦節比翼列ディフューザ / 多目的最適化 / 数値流体力学 / 遠心圧縮機 / 人工神経回路網 / 遺伝的アルゴリズム / 流体工学 / ニューラルネットワーク |
|---|
| Outline of Final Research Achievements |
A technical issue for turbomachinery is not only for operation of a design condition but also for operation of an off-design condition. Classical design method is only focused on the flow condition at the design condition with reducing a loss based on flow separation. However, if it considered the operation at off-design condition, it is impossible to escape from generation of flow separation, as a result, unstable flow is observed such as rotating stall and surge. The objective of this study is to propose a design which has a function of self-repair of the flow using secondary flow at off-design condition. Multi-point multi-objective optimization system is applied for global search. Genetic algorithms with a meta-model of neural network is effective to reduce the computational cost. As a result, a novel design of a recirculation flow type casing treatment and a low solidity diffuser are found for the flow range enhancement with an effect of self-improvement by secondary flow.
|
| Academic Significance and Societal Importance of the Research Achievements |
人工知能の利用は流体機械の設計にも有効であり，従来検討していなかった形状まで全方位的に探索できるようになった．本課題でも，流体機械の非設計点における設計を行う際に，人工知能を利用することで，効率的な形状探索を行うことができた．ただし，コンピュータにどのようなデータを与えて学習させ，何を期待するかを明確に示さなければ，コンピュータの提案する形状は最適なものとはならない．本課題では，二次流れの積極的利用というアイデアを具現化するために，最適形状を探索させ，従来に成しえなかった性能を得ることができた．流体機械の性能改善という具体的な提案だけでなく，人工知能の利用方法について指針を示すことができた．
|
