A Method for Predicting Aerodynamic Performance of Airfoils Based on Selective Fusion Neural Networks

Prediction and Plotting of Aerodynamic Performance of Airfoils

Input flight conditions and CST parameters to rapidly generate airfoil geometry and predict Cl/Cd/Cm. Supports 5-fold cross-validation for integrated inference, with results including L/D for auxiliary evaluation.
Access the Page

Proxy Model Performance

Review model metrics, including comparison tests, prediction scatter plots, and residual histograms (Not open yet).
Access the Page
1695
airfoils
163363
datasets
10
inputs
Click to view model documentation (Not open yet).
View Model
Contact Information: Changchun Institute of Applied Chemistry, Chinese Academy of Sciences · YanchiLi · yc.li@ciac.ac.cn · 2025.9