Airfoil profile flow field prediction network training method, airfoil profile flow field prediction network, airfoil profile flow field prediction method and medium

• Main Authors: JIANG YUBIAO, GUO LINLIANG, ZHANG JUN, JIANG HAILIN, LIANG LEI, YU RUIZHE
• Format: Patent
• Language: Chinese; English
• Published: 16.07.2024
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS

The invention relates to an airfoil flow field prediction network training method, a network, a prediction method and a medium, and the method comprises the steps: obtaining a plurality of airfoil samples, calculating flow field data under different attack angles, Reynolds numbers and Mach numbers, and taking the airfoil samples containing the flow field data as real flow field images; and taking a real flow field prediction result corresponding to the real flow field image as a label to obtain a training sample. Extracting coordinates of points in the flow field point clouds of the training samples, obtaining a first feature map based on the coordinates of the airfoil points, and obtaining a second feature map of the coordinates of the points in the flow field point clouds and the corresponding Mach number, Reynolds number and attack angle; acquiring a third feature map including the Mach number, the Reynolds number and the attack angle, and a fourth feature map including coordinates of all points in the flo