A weighted hybrid model for unsteady nonlinear aerodynamics

Abstract

Based on experimental data from the large amplitude oscillation experiment conducted with two degrees of freedom, this work studies and compares the ability of Polynomial Regression, LS-SVM and RBF network to describe the characteristics of unsteady nonlinear aerodynamics. This work also develops a hybrid model for use in unsteady nonlinear aerodynamics based on the standard boosting method. The results indicate that the forecast results and actual data are in good agreement using the method, thus demonstrating that these methods can effectively model highly nonlinear aerodynamics. The results also indicate that the hybrid model has a better effect compared to other methods.