顾及时间相关噪声的单频GPS精密单点定位

扩展Kalman滤波(EKF)常常被用于单频GPS精密单点定位。Kalman滤波的前提假设之一是观测噪声为白噪声,即时间不相关,这在实践中往往不能满足。因为单频GPS观测值中包含有很难被完全消除的电离层、对流层等大气折射误差,以及多路径影响误差。这些误差在时间上是相关的,严重地影响了滤波解的精度和收敛时间。这里提出一种顾及时间相关噪声的Kalman滤波,在移动窗口内,利用核估计预测时间相关噪声的系统部分,进而实时修正当前历元的观测值和观测向量的协方差矩阵。该方法还有一个明显的优点就是,在滤波过程中不需要对时间相关误差做任何假设。最后通过一个实测算例验证了该算法的适用性。

Consideration of Time-Correlated Errors in a Kalman Filter Applicable to Single-Frequency GPS Precise Point Positioning

Extended Kalman filter(EKF) is adapted to single-frequency GPS precise point positioning,the filter can perform optimally only when its dynamic model and stochastic model are correctly defined.In particular, the assumption that the measurement errors are uncorrelated over time is often unjustified.An excellent example is with single-frequency GPS measurements that are corrupted by various time-correlated errors such as orbital errors,clock errors,atmospheric errors and multipath,which actually degrade posit...