双频精密单点定位观测模型研究

Research on observation models for dual-frequency precise point positioning

精密单点定位是目前G N SS精密定位领域的研究热点之一,在过去的十几年里,研究人员提出了多种精密单点定位观测模型,包括传统的非差模型、UofC模型、星间差分模型和历元差分模型。这些模型都是基于双频伪距和相位观测量,通过线性变换得到的。因此,从理论上说,各种模型应该等价。本文从原始观测量出发,推导了上述几种精密单点定位观测模型和随机模型,采用实测的静态和动态数据,比较了不同模型的定位精度和收敛速度。研究发现:传统非差模型在精度和收敛速度上都最好;UofC模型在精度和收敛速度上与传统模型相当,但是计算量较大;星间差分模型的定位精度与前二者相当,收敛速度慢,收敛之前定位精度较差;历元差分模型的定位效果最差,且收敛速度慢,但计算量较小。

Precise point positioning is a popular research topic in GNSS precise positioning field.Several models have been proposed for dual-frequency precise point positioning in the past ten years,including traditional un-difference(UD)model,UofC model,between-satellite single difference(SD)model and time difference(TD)model.All of these models are derived from the original dual-frequency pseudorange and phase observations by linear transformation.Therefore these models should be equivalent in theory.In this paper,different observation models and their stochastic models are derived and analyzed in detail.Both static and dynamic data were collected to compare the positioning precision and convergence time of the four models.The results show that UD model delivers the best performance in terms of precision and convergence;UofC model equals the traditional model in precision and convergence,but it costs much more time to compute;SD model is comparable with the previous two models in positioning precision,but it has a slow convergence and the precision before convergence is not good;TD model gives the worst performance in precision,with a slow convergence and a small amount of calculation.