| Autor: |
Ze WANG, Shufang SONG, Xu WANG, Weiwei ZHANG |
| Jazyk: |
čínština |
| Rok vydání: |
2024 |
| Předmět: |
|
| Zdroj: |
气体物理, Vol 9, Iss 4, Pp 39-55 (2024) |
| Druh dokumentu: |
article |
| ISSN: |
2096-1642 |
| DOI: |
10.19527/j.cnki.2096-1642.1068 |
| Popis: |
The accurate prediction of aerothermal loads is the basis to guide hypersonic vehicle design. Under the background that classical aerothermal prediction methods are more and more difficult to meet the demand of efficient and accurate aerothermal prediction in engineering, data-driven aerothermal modeling prediction methods have gradually become a new paradigm of aerothermal prediction in recent years. Firstly, the relationship between the data-driven aerothermal modeling prediction method and the classical aerothermal prediction method was described. Then, from the modeling idea, the data-driven aerothermal modeling prediction methods were summarized into three categories: The dimensionality reduction modeling method of feature space, pointwise modeling method and physical information embedding modeling method were introduced and analyzed in detail. It is found that the data-driven aerothermal modeling prediction method is not only more accurate than the engineering algorithm, but also can effectively reduce the workload of test measurement and numerical calculation when combined with the sampling method, and the model given is more efficient and concise. Finally, the development trend of data-driven aerothermal modeling prediction methods was prospected. It is pointed out that the deep combination of data-driven technology and classical aerothermal prediction methods, aerothermal physical information embedding modeling methods and aerothermal prediction big models will be the key points of future research. |
| Databáze: |
Directory of Open Access Journals |
| Externí odkaz: |
|
