基于卡尔曼滤波重建的GPS点位修正

GPS point correction based on Kalman filter reconstruction

由于单点定位的结果受卫星星历误差、卫星钟误差以及卫星信号传播过程中大气延迟误差的影响较为显著,因此解算出的定位结果在真值附近上下浮动。文中采用ARMA模型建立卡尔曼滤波的观测方程和状态方程,并对定位结果进行滤波;采用一次滤波后的坐标值作为初值,建立ARMA模型并二次滤波。实验表明,滤波有效防止了定位结果偏差过大情况的发生,使滤波收敛值与准确值最大偏差不超过3cm,表明采用一次滤波后的坐标值建立的模型更为合理,从而为单点定位结果的时间序列模型的建立提供一种新方法。

The single point positioning results are affected markedly by the satellite ephemeris error,the satellite clock error and the atmospheric delay error.So the calculated positioning results fluctuate around the truth value.ARMA model is used to establish the Kalman filter observation equation and state equation,and filtere the positioning results.The coordinate values are used as the first time filtering which is regard as initial value to establish ARMA model and carry out second time filtering.The experiments show that it can effectively prevent the big deviation of positioning results and the maximum deviation between filtering convergence value and accurate value is less than 3cm.It shows that the model established by first time filtering coordinate value is more reasonable.So a new method for the establishment of single point positioning time-series model is provided.