Parameter Fitting and Prediction for Hypersonic Glide Vehicles Based on Characteristic Extraction and VARMA Model

The parameter fitting and prediction is fundamental to the defense and interception for hypersonic glide vehicle (HGV). In this article, we propose a parameter fitting and prediction method for HGV based on characteristic extraction and vector auto-regressive moving average (VARMA) model. First, the maneuver mode of HGV is explored from the perspective of acceleration and the aerodynamic lift and drag parameters, and the bank angle are selected as the predicted parameters on the basis of maneuver mode analysis. Then, the nonstationary characteristic extraction is performed by discrete Fourier transform and difference. Finally, the parameter fitting and prediction is carried out based on the VARMA model. The model orders are determined utilizing the minimum description length criterion and the minimum root mean square error of predicted trajectory. The coefficient matrices estimation is converted to a convex optimization problem, and thus, a unique and accurate solution is ensured. The proposed method can efficiently deal with the nonstationary characteristics and unknown noise. Simulations demonstrate that the proposed fitting and prediction method can achieve better prediction accuracy than the existing methods.