Abstract: To provide priori information of situation and threat assessment and interception defense for hypersonic gliding reentry vehicle, a multi-level recursive trajectory prediction method is proposed. Referring to the theory of multi-level recursive, the method compensates for the unknown parameters of prediction model stochastically. Accordingly, the trajectory prediction problem is decomposed into prediction of aerodynamic parameters and prediction of the target trajectory. The dynamic model is carried out by augmenting the aerodynamic parameters into state vector, and the mix estimation is done for the aerodynamic parameters and model error. The time series forecasting is performed using the estimation of aerodynamic parameters. Based on the prediction of aerodynamic parameters, the dynamic integral equation is used to predict the target trajectory. Two scenarios for different flight modes are designed to analyze the influence of tracking and prediction time on prediction accuracy. Simulations show that the algorithm has a stable and reliable capability for trajectory prediction.

Key words: hypersonic vehicle, reentry glide, trajectory prediction, dynamic model, multi-level recursive, time series forecasting, model compensation

摘要： 为给高超声速滑翔目标态势与威胁评估、拦截防御等提供先验知识,提出一种多层递阶轨迹预测方法。该方法借鉴多层递阶预测理论对预测模型进行随机补偿,将轨迹预测问题分解成气动参数和模型误差的混合预测以及在此基础上对目标轨迹的预测。方法首先利用气动参数增广状态向量进行动力学建模,对气动参数和模型误差进行混合估计,根据参数估计值进行时间序列预测。然后,在参数预测的基础上,利用动力学模型积分预测目标轨迹。仿真设计了2种有规律的飞行模式仿真场景,分析跟踪与预测时间对预测精度的影响,结果表明算法具有稳定可靠的轨迹预测能力。

关键词: 高超声速飞行器, 再入滑翔, 轨迹预测, 动力学模型, 多层递阶, 时间序列预测, 模型补偿