Fast sparse flow field prediction around airfoils via multi-head perceptron based deep learning architecture
Based on the UIUC database, the relevant airfoil parameterization work is carried out:UIUC Airfoil Data Site (illinois.edu)
1、airfoil_interpolate.py
used to interpolate scattered airfoil coordinates into 70 uniform coordinate data
2、First run get_xy_coordinate.py
get x,y coordinates of the interpolated different airfoils
3、Run get_x_coordinate_averge.py
get a fixed x-coordinate along the airfoil chord length
4、Because the image drawn by Python is in ARGB format, it is necessary to run convertGray2.py
convert it to a single-channel grayscale image
5、get_train_files.py
get the deep learning model training file
6、Run the CNN_train.py
file to train the model
7、Run the CNN_test.py
file to train the model
8、A similar file that begins with mlp represents code related to MLP flow field prediction
9、A similar file that begins with MHP represents code related to MHP flow field prediction