卡尔曼滤波卫星授时的仿真技术

为了提高无线导航卫星系统授时精度,提出了采用卡尔曼滤波对用户时钟钟差、钟漂及老化率进行估计,并预测出下一采样时刻用户时钟与系统时的差值的闭环授时方法,设计了卡尔曼滤波器.分别对单星和多星的授时问题进行了讨论,理论证明了采用多颗卫星授时可提高授时精度,并给出了克拉默罗界.利用STK软件实时产生卫星三维位置和速度,建立了用户时钟仿真模型和观测误差仿真模型,对所提出的方法在单星和多星的情况下进行了仿真.结果表明,在噪声较大情况下仍能保持较高的授时精度.

Simulation technology for Kalman filtering in satellite time transfer

In order to impro ve the accuracy o f time t ransfer of RNSS( radio navigat ion satellite sy stem) , anovel closed lo op metho d of Kalman f il tering was proposed to est imate the clock o ff set , clock shift ing andclo ck aging rate, so as to predict the clo ck bias betw een user t ime and system t ime at the next samplingt ime, and the design of Kalman f iltering w as g iven. Moreov er , the situat ion of single satellite and mult iplesatellites w ere both discussed, theo ret ically proving that the performance of time t ransfer can be impro vedby using mult iple satellites, and Cramer-Rao bound was presented as w ell . Finally , the threee-dimensionpo sit ion and v elo city of each satellite w ere gener ated by ST K. User clo ck model and observ at ion err ormodel w ere set up to simulate the method in the co ndition of single satel lite and mult iple satell ites. T he resultindicates that the accuracy of t ime t ransfer can be kept even under heavy noise situat ion.