基于强跟踪滤波的天基测角卫星定位算法

在天基仅测角卫星定轨扩展卡尔曼算法(EKF)中，EKF关于模型不确定性的鲁棒性较差，造成滤波器估计不准，甚至发散。提出一种带次优渐消因子的扩展卡尔曼滤波(SFEKF)定轨方法，利用渐消因子对过去的数据进行渐消，减小了模型的截断误差，提高了滤波的收敛速度和对卫星的跟踪能力。仿真结果表明，与EKF定轨方法相比，SFEKF方法具有更高的定轨精度、稳定性和收敛速度。

The Orbit Determination of Non-cooperative Satellite used Strong Tracking Filter by Applying Space-based Bearing-only Measurement

During the determination of the orbit of a non-cooperative satellite in space by using Extended Kalman filter, the robustness of model uncertainty is poor, and may cause divergence of EKF. In this paper, an orbit determination using suboptimal fading EKF is proposed. Suboptimal fading factor is using to fading the old observation, in order to decrease the truncation error of model and improving the stability and tracking ability. The mathematics simulation results show that the location performance of SFEKF is superior. Compared with the EKF, the SFEKF can obtained a higher accuracy of estimation and more stable.