网络RTK基站数据处理随机模型验前估算

Prior estimation of stochastic model in network-RTK data processing

针对网络RTK基站数据处理中,基于经验的随机模型无法在异常情况下反映实际观测值的精度,以及基于基站数据处理验后残差的随机模型计算量大、处理过程复杂等问题,提出一种利用单站观测值历元间差分模型的估计残差对基站数据处理随机模型进行验前估算的方法。该方法计算量小、处理过程简单,所建立出的随机模型能较为准确地反映实际观测值的误差统计特性。实验结果表明,该方法相比于等精度模型及高度角模型在定位精度方面有3%~5%的提升,在收敛速度方面有15%以上的提升。

Aiming at the problems that the stochastic model based on experience cannot reflect the actual accuracy of the observation in abnormal conditions, and the stochastic model based on the posterior residual of data processing of the base sta- tion suffers intensive computation and complex process in network RTK data processing, the paper proposed a prior estimated method of the stochastic model utilizing the residuals estimated by epoch-difference model of the single station, which involving small amount of calculation and a simple process, and reflecting the statistical characteristics of errors of the actual measure- ments. Experimental result showed that compared with equivalent-accuracy models and elevation models, an improvement of 3% ~5% in positioning accuracy, and of 15% in convergence rate could be obtained by the proposed method.