一种利用星间仅测向数据的滤波定轨算法

基于星间方向观测数据实时确定目标星轨道，一般扩展Kalman滤波算法是建立角度直接观测量与状态量的线性关系作为观测方程。由于缺少距离测量值，仅包含角度信息的观测方程对状态变量的约束较弱，导致滤波算法收敛速度较慢。〖JP3〗将用于初轨计算的改进拉普拉斯方程作为观测方程，这样,观测方程不仅包含了两星的几何信息，还充分考虑了目标星的轨道约束。仿真结果表明，在同样测量精度下，该算法的收敛性能明显优于一般滤波算法。

Orbit determination filter using satellite-to-satellite bearings-only measurement

The observation equations of the current extended Kalman filter (EKF) describe the linear relationship between state vector and bearings measurement, which only exerts weak constraint on state vectors without the support of relative range measurement. To solve the problem, the observation equations are rebuilt by the improved Laplace method in a creative way, which considers both the geometry and orbit restrictions. Simulation results show that the convergent performance is dramatically improved by this method compared with the current EKF with the same measurement precision.