空间碎片激光测距中轨道偏差的非线性传播OA北大核心CSTPCD

In laser ranging of space debris,the success rate is often low due to significant orbit prediction errors.To address this,a method that combines the unscented transfor-mation technique with the Gaussian Mixture Model to propagate the orbital uncertainty of space debris is proposed.This approach yields the debris's state moments(generally mean and covariance matrix)or probability density function(PDF)at their terminal time,thereby providing a robust foundation for laser ranging.First,typical orbits in current laser rang-ing of space debris are selected to establish initial orbit uncertainty.Then,the unscented transformation method is utilized to nonlinearly predict the mean and covariance matrix of this uncertainty.Subsequently,the Gaussian Mixture Model is employed to approximate the probability density function of the uncertainty,and the results are converted into the station coordinate system centered on the station.The outcomes align with those derived from the Monte Carlo method:the absolute error of predicted range remains within 20 m,while the errors in azimuth and elevation angles do not surpass 2 arc seconds.Furthermore,the predicted probability density function curve closely mirrors the Monte Carlo simulation results,and the computational time efficiency is enhanced by over one hundred times.This demonstrates that the method not only offers high-precision uncertainty prediction but al-so furnishes a reliable foundation for subsequent laser ranging.Additionally,it effectively mitigates the substantial computational demands associated with Monte Carlo simulations.