Application of Shallow Learning in Aerodynamic Prediction of Wind Tunnel Test

Wind tunnel test is an effective means to obtain aerodynamic performance of aircraft. Due to the limitation of test conditions and cost, the aerodynamic data obtained directly from the test are often small and incomplete, so it is meaningful to build an accurate aerodynamic prediction model to meet the engineering application. In this paper, the traditional interpolation method, Kriging method, Support Vector Machine method and RBF neural network method are used to aerodynamic prediction of a transport aircraft and a small aspect ratio aircraft, and the advantages and shortcomings of these methods are compared and analyzed. The results show that Kriging method is better than traditional interpolation method in wind tunnel test aerodynamic prediction, and has a good application prospect.