PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
March 3, 2026Scientia Sinica Technologica0 citationsOpen Access

基于多尺度几何先验神经算子方法的飞行器气动性能预测

WTWeilong TanXGXiaohui GuSDShiran Dai

Key Points

  • The method effectively maps point clouds to local refined representations, enhancing geometric feature capture.
  • Correlation coefficients above 0.95 for wing-body configurations validate the method's accuracy in aerodynamic predictions.
  • Adaptive sampling controls point cloud density based on regional importance, optimizing geometric data input.
  • This approach may enable rapid assessments in complex configurations, proving robust across diverse aerodynamic scenarios.

Abstract

针对传统气动仿真重复迭代多, 难以支撑整机级快速设计评估的问题, 本文提出了一种基于多尺度几何先验神经算子的新颖方法. 利用基于网格感知的几何重映射 (Grid Aware Geometric Remapping, GAGR) 可将原始点云映射为信息分布集中的局部加密表示, 从而在有限点数要求下更有效地承载关键几何特征; 空间约束下的适应性采样机制 (Adaptive Sampling under Spatial Constraints, ASSC), 通过体素划分与重构局部点集以有效调控输入几何中不同区域的点云密度, 形成基于区域重要性的自适应密度调控并实现具备合理空间分布的点云几何. 最后, 引入几何信息神经算子模型, 通过在潜在规则网格空间中利用局部几何邻接关系与傅里叶算子结构, 实现从外形几何到气动性能的端到端映射. 在翼身融合构型上的实验表明, 本方法可准确的重现表面压力系数的分布特征, 预测结果的相关性系数可达0.95以上; 在自建的背部进气式布局构型数据集上的相关性系数可达0.91以上, 结果验证了所提方法在复杂构型场景下的鲁棒性, 为航空飞行器气动性能快速评估任务提供了高效可行的方案.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Tan et al. (2026) studied this question.

synapsesocial.com/papers/69a76043c6e9836116a2cd1fhttps://doi.org/10.1360/sst-2025-0384
Ask AI
Helpful
Bookmark
Share
View Full Paper

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Machine learning-assisted sparse observation assimilation for real-time aerodynamic field perception2024 · 18 citations
  2. 2Parameter-Efficient Fine-Tuning in Spectral Domain for Point Cloud Learning2025 · 19 citations
  3. 3Development and deployment of data-driven turbulence model for three-dimensional complex configurations2024 · 6 citations
  4. 4Data-driven surrogate model for aerodynamic design using separable shape tensor method2024 · 22 citations
  5. 5Multi-objective optimization of the streamlined head of high-speed trains based on the Kriging model2012 · 36 citations