利用固定模糊度的动态精密单点定位算法

Dynamic Precise Point Positioning Algorithm Based on Fixing Ambiguities

首先,采用序贯最小二乘法计算无电离层组合观测值的模糊度;然后,固定宽巷组合模糊度;再固定窄巷组合模糊度;最后,得到无电离层组合观测值的模糊度最终解。谱密度的取值影响状态参数预测值的协方差矩阵元素的大小,因此,采用自适应滤波进行处理。利用机载GPS数据进行验证,结果表明,与其他方案相比,利用固定模糊度的自适应滤波加快了收敛速度,提高了动态精密单点定位的解算精度;无论谱密度取何值,自适应滤波都能够得到较稳定的解。

The convergence speed of Kalman filter is slow and the accuracy of initialization is low if the ambiguities are taken parameters estimated in dynamic precise point positioning. The accuracy of Kalman filter will degrade if the spectral density of the dynamic model is not accurate. So fixing single difference ambiguities are used to improve the speed of convergence. Firstly, the ambiguities of ionosphere-free observations are estimated with the sequential least squares, and the ambiguities of wide lane are fixed. Then, the ambiguities of narrow lane are fixed. At last, the ambiguities of ionosphere-free observations are fixed. For errors of the spectral density affect covariance matrix of the predicted state vector, adaptive filtering is used to control outliers. Adaptive factor on the basis of current information can adjust the scale between covariance matrix of predicted state and covariance matrix of observations noise, so that the contribution of dynamic state to results of Kalman filter is more efficient. In the processing of data of air, the adaptive Kalman filter in which single difference ambiguities fixing are used, can improve the speed of convergence and accuracy of positioning. The stability of adaptive Kalman filter is better when the spectral density has different numerical value.