Airfoil profile surface aerodynamic parameter prediction method based on point cloud network

• Main Authors: LIU RUOXUAN, SONG JUZHENG, MENG XINYI, WANG ZHE, XUE CHANGBAO
• Format: Patent
• Language: Chinese; English
• Published: 20.08.2024
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS

An airfoil profile surface aerodynamic parameter prediction method based on a point cloud network comprises the following steps: selecting an initial airfoil profile, parameterizing the initial airfoil profile by using a geometric parameterization method, forming a random airfoil profile library, obtaining surface aerodynamic parameters of each airfoil profile in the random airfoil profile library, establishing a data set, training the point cloud network by using the training data set, and predicting the aerodynamic parameters of the airfoil profile. The test data set is used for testing and verifying the prediction performance of the point cloud network, coordinates of the airfoil surface coordinate points serve as input, the trained point cloud network is input, output values of the network serve as prediction values of the airfoil surface aerodynamic parameters, and prediction of the airfoil surface aerodynamic parameters is completed. 一种基于点云网络的翼型表面气动参数预测方法，选取一个初始翼型，应用几何参数化方法对初始翼型进行参数化，形成随机翼型库，获取随机翼型库中各个翼