Flutter boundary prediction under continuous variable speed based on time series models

Based on the time series model, the modal parameters of the continuous variable flutter test response signal are identified, and a set of flutter boundary prediction methods suitable for turbulent excitation is developed in this paper. In order to ensure the accuracy of the method, a modal parameter identification method is used to analyze the traditional autoregressive model (AR) and the time-varying autoregressive model (TVAR), and compare the accuracy of the two models. Finally, the method is applied to the flutter boundary prediction of turbulence signals. The prediction method combines the time series model with the stability criterion, constructs the stability parameters of the response signal, and the prediction results of flutter critical velocity are obtained by fitting and extrapolation. The numerical example shows the analysis results of the two models and proves the feasibility and effectiveness of the method. Finally, this method is used to predict the flutter boundary of low-speed wind tunnel test data, and the prediction error is less than 5%.