仅测角卫星跟踪的扩展卡尔曼粒子滤波算法

Satellite bearings-only tracking using extended Kalman partcle filter

卫星被动跟踪技术对于空间监视系统非常重要,为提高非合作跟踪模式下空间低轨卫星(LEO)的跟踪精度,提出一种基于扩展卡尔曼粒子滤波(EKPF)的仅测角卫星被动跟踪算法。首先,在卫星二体运动下,以地球惯性直角坐标系为基础,对卫星运动状态方程的二阶线性化、状态方程和测量方程的雅可比矩阵进行了详细推导。其次,利用二阶扩展卡尔曼滤波(EKF)产生粒子滤波(PF)的建议分布,并将EKPF应用到低轨卫星跟踪。最后,通过STK6.0仿真场景产生的数据对EKPF算法和扩展卡尔曼滤波算法的跟踪性能进行了对比分析和验证。仿真结果表明,该算法可以实现单颗高轨卫星通过仅测角对圆轨道的低轨卫星进行跟踪,且跟踪精度优于EKF算法。

Satellite passive tracking techniques have great significance in space surveillance system. To improve low earth orbit（LEO） satellite tracking precision in non-cooperative tracking mode, a new passive extended Kalman particle filter （EKPF） algorithm using bearings-only measurements in earth centered inertial （ECI） coordinate was proposed. Firstly, the approximated second order lineafized satellite state propagation equation, Jacobian matrix of state transfer equation and measurement equation were derived in the condition of satellite two-body movement. Secondly, the second order extended Kalman filter（EKF） was used to generate the proposal distribution within the particle filter （PF） framework, and the second order EKPF algorithm was used to track LEO satellite. The tracking performance of EKPF algorithm was compared with extended Kalman filter through simulated data from STK6.0. Simulation results indicate that it is possible for a high earth orbit satellite to track low earth circular orbit satellite passively using bearings-only measurements, and the tracking precision is higher than EKF algorithm.