|
CSME 2022/12
Volume 43 No.6
:
509-520
|
|
Prediction of Transonic and Subsonic Wind Tunnel Aerodynamic Data by Neural Networks
Jun-Kai Ouyanga, Yi-Ting Liaoa, Ying Lib and Wei-Hsiang Wangb
aDepartment of Aerodynamics, Aeronautical Systems Research Division, National Chung-Shan Institute of Science and Technology, Taichung, Taiwan 407104, ROC
bDepartment of Mechanical Engineering, National Chung-Hsing University, Taichung, Taiwan 40227, ROC
|
Abstract:
This study aims to build the backpropagationneural network model to predict wind tunnel aerodynamicdata. The experiments were performed to obtainthe pressure fluctuation on compressible cavity flowsin transonic wind tunnel and used different sweepangle delta wing model to obtain the aerodynamic datain subsonic wind tunnel, respectively. The experimentsdata is used as the training parameter for neuralnetwork to decide the neural network structure, tuningthe adjustable hidden layers and neuron numberparameters of the neural network. The Levenberg–Marquardt (LM) technique is adopted as the weightingtraining algorithm to minimum the cost function. Thisarticle have established the neural network model toprovide good agreement with experiments result. Byusing neural network technique, the wind tunnel testefficiency and aerodynamic data analysis can besignificantly improved.
|
|
Keywords: neural networks, wind tunnel, cavity flow, delta wing
|
Download PDF
|
|
*Corresponding author; e-mail:
|
©
2022
CSME , ISSN 0257-9731
|