A Modelling Method of High-altitude Wind Uncertainty based on Non-stationary Gaussian Process
Modelling methods of the high-altitude wind uncertainty have significant influence on aerodynamic loads and the overall performance optimization of launch vehicles.The commonly used methods in engineering do not consider the correlation between wind speeds at different altitudes,and the simulated wind samples are inconsistent with actual wind characteristics,therefore disable to accurately predict the aerodynamic loads.A modelling method using U/V wind speed components as non-stationary Gaussian processes about the height is proposed.The statistical distribution parameters of massive actual historical wind samples are analyzed,and the expansion optimal linear estimation(EOLE)method is adopted to generate wind samples with the same distribution as actual wind samples.Results indicate that the proposed method can simulate the actual wind profile more precisely than traditional methods.The simulation program of real launch vehicle is used to calculate the wind loads,and the results show that the wind loads predicted by the proposed method is far more accurate than traditional methods,which is of great significance to improve the launch efficiency and launch probability of launch vehicles.
NETL
NSTL
万方数据
