一种低轨卫星轨道预报算法

在基于EMBET（Error Model Best Estimation of Trajectory，误差模型最佳弹道估计）的雷达自校准处理过程中，需要基于给定的低轨卫星初始状态完成轨道预报，得到与雷达测量序列对应的卫星位置序列，进而通过迭代收敛过程得到误差模型参数估计。针对低轨卫星轨道迭代预报的需要，提出了一种边积分边插值的轨道预报方法。即将低轨卫星轨道积分计算过程的加速度计算结果直接应用到三弯矩插值，省略了三弯矩插值中节点矩的计算过程，同时将积分步长以及插值区间步长设置为1 s，减少了积分及插值中乘法和除法的计算次数，进一步提高了计算效率。经仿真计算表明，该方法在保证轨道预报精度的情况下，计算效率比传统低轨卫星轨道预报方法提高了约20倍，有效提高了雷达自校准处理的时效性。

A Novel Algorithm for Prediction of LEO Satellite Orbits

Prediction of Low Earth Orbit(LEO) satellite orbit based on its initial state is needed in processing of radar self-calibration that estimates some parameters by the Error Model Best Estimation of Trajectory (EMBET) method to get a time-series of satellite positions. To get a precise result, the process must be done many times. A novel algorithm hybridized with integration and interpolation is proposed for the processing, and it improves the computational efficiency and flexibility without changing the accuracy of results. It omits the computation of moment parameters in three-curve-moments interpolation by getting them from the integration procedure. Meanwhile, it reduces the computation times of multiplication and division in integration and interpolation by setting the step to one second, thus further increasing computational efficiency. Simulation results show that the method increased the computational accuracy of LEO satellite orbit prediction by about 20 times compared with traditional LEO satellite orbit prediction methods, and it effectively improved the effectiveness of radar self-calibration processing.