摘要
闪烁噪声下的卫星被动测角定位问题是一个非线性非高斯的目标跟踪问题,传统的滤波算法很难保证滤波精度。为提高跟踪算法的鲁棒性,在容积卡尔曼滤波(CKF)中引入渐消因子,提出了带渐消因子的强跟踪容积卡尔曼滤波。算法采用三阶容积数值积分原理代替强跟踪滤波(STF)中的Jacobian矩阵的计算,由得到的统计量计算渐消因子,扩展了STF的适用范围,提高了CKF的鲁棒性。仿真结果表明,在闪烁噪声的影响下,采用ST-CKF的卫星被动测角定位算法有效地提高了滤波精度和稳定性。
The orbit determination under glint noise is a nonlinear and non-Gaussian problem, which means traditional Kalman iiher cannot guarantee the tracking precision. Fading factor was introduced in Cubature Kalman Filter （CKF） by putting forward Strong Tracking Cubature Kalman Filter （ST-CKF） with fading factor. The algorithm calculated the fading factor by means of numerical integration based eubature rule instead of caculating Jacobian matrix. The current observation has more effect to the state value by means of the fading factor, which improves the stability of CKF. The simulation results indicate that higher accuracy and stability of state estimation is obtained in comparison with CKF and ST-CKF under glint
出处
《计算机应用》
CSCD
北大核心
2014年第A01期345-348,共4页
journal of Computer Applications
基金
国防十二五预研项目(51322010601)
关键词
非合作
卫星定位
闪烁噪声
容积卡尔曼滤波
渐消因子
non-cooperation
orbit determination
glint noise
Cubature Kalman Filter （CKF）
fading factor
