【作者】 赵磊；

【导师】 张守建；

【作者基本信息】 武汉大学 ， 大地测量学与测量工程， 2017， 硕士

【摘要】 自2008年以来,许多国内外的学者研究了全球导航卫星系统(Global Navigation Satellite System,GNSS)精密单点定位(Precise Point Positioning,PPP)技术中的模糊度固定方法,指出通过固定模糊度,PPP的收敛时间和精度都能得到显著改善,而且多卫星系统的组合效果更明显。本文在导师提供的软件平台上研究了 PPP固定解技术的实现。浮点PPP模型中的模糊度偏差参数不具有整数特性的原因是该参数除了包含整数模糊度外,还包含卫星端和接收机端的未校正的相位偏差(Uncalibrated Phase Delays,UPD),这些小数偏差都和硬件相关。PPP模糊度固定解技术的实现即是要从浮点模糊度参数中分离出这些偏差的小数部分(Fractional Cycle Biases,FCB),恢复模糊度的整数特性,从而实现模糊度的固定。许多学者提出的固定解技术的本质区别就在于分离小数偏差的方式不同。一类方法是把浮点模糊度参数在卫星间进行单差,消除公共的接收机端的UPD,而计算卫星间差分后的UPD,并将之用于单差浮点模糊度的改正;另一类方法不进行作差,而是通过适当地合并观测方程中的参数,使伪距和载波相位观测值模型中包含不同的接收机和卫星钟差参数,直接估计卫星和接收机端的UPD,进而用来改正用户端浮点模糊度。本文实现的分离方法和后一类的思路相近,不进行作差,但是对伪距和载波相位观测值的建模仍使用同一套钟差参数,逐个历元估计卫星和接收机端的UPD;值得注意的是,这样的处理方式会使得浮点模糊度参数还会包含码间偏差(Differential Code Biases,DCB)。所以,为了使用计算出的UPD,需要保证用户端和服务器端数学模型的一致。在实验部分,本文首先测试了提出的一种新的组合的周跳探测方法以应对电离层的扰动,接着以相对定位的方式展示了一种比较高效的部分模糊度的固定策略,并也将之用于了后续的PPP固定解实验中。然后,实验部分重点测试了实现的PPP固定解技术。首先,测试了 GPS的PPP固定解效果,根据计算的5天的统计结果,静态PPP浮点解的精度在水平方向为lcm左右,天顶方向优于5cm;静态的PPP固定解的精度在水平和天顶方向都有提升,但不明显。其次,动态PPP浮点解的精度在水平方向为2cm左右,天顶方向为5cm左右;而相应的固定解结果,在水平方向提升至1cm左右,在天顶方向优于4cm。从收敛时间来看,静态模式下,PPP固定解有所改善,但也不明显;而动态模式下,PPP固定解的收敛时间明显缩短,比如测试的3个用户测站的PPP收敛时间,都至少缩短了 50%。此外。本文还选取了 MGEX(Muti-GNSS EXperiment)台站的观测数据,利用提出的UPD估计模型,分别又估计了 GPS和Galileo系统UPD,并测试了两个系统的PPP固定解结果。根据结果来看,3个用户测站GPS的PPP固定解的收敛时间明显短于浮点解;收敛后的静态PPP固定解比较稳定,但是动态PPP固定解不够稳定,可能是由于所计算的卫星UPD精度不够造成的。对于Galileo系统,由于可观测到的卫星数量较少,Galileo系统的动态浮点解很难收敛,因此,实验部分只测试分析了 Galileo系统的静态PPP固定解效果。结果表明,计算的Galileo系统卫星的窄巷UPD的精度比较差,并且直接影响到了静态固定解的定位精度。这说明本文提出的模型在估计窄巷UPD时还有待进一步改善。另一方面,多系统组合的PPP固定解技术仍有待研究。更多还原

【Abstract】 Since 2008,the issue of Ambiguity Resolution(AR)in Precise Point Positioning(PPP)technology has been a very popular research point among the Global Navigation Satellite System(GNSS)community.Many studies in home and abroad show that fixing the ambiguous phase biases in PPP solutions can significantly shorten the convergence time and improve the positioning accuracy.Based on the software platform provided by my tutor,we implemented the PPP AR test.The fact that the Uncalibrated Phase Delays(UPD)exist in both satellite and receiver end makes the ambiguity biases fractional.These delays are all instrumentally dependent.Implementation of PPP AR is essentially to separate the fractional part of UPD from phase biases parameter and to recover its integer property.Hence,AR strategies proposed by many researchers are different in the way of separation.For example,one of them fixes the between-satellite single differenced ambiguity by correcting the related single differenced UPD,in which UPD from receiver has been cancelled duo to the operation of difference.Another one fixes the un-differenced ambiguity directly by estimating the instrumental UPDs with so called"decoupled clock model".In this effort,our approach is similar to the latter one,but we still use common clock error parameters in the code and phase measurement equation,and explicitly estimate the UPD both from satellite and receiver end epoch by epoch.It should be noted that the UPD parameter in our approach also contains the code instrumental biases.Therefore,it is necessary to make sure the mathematic-model consistency between server and user end.In the experimental stage,we firstly tested a combined cycle slip detection strategy to handle the active ionospheric phenomenon.Then we investigated a proposed partial AR strategy,which was also used in the following PPP AR tests.Next,GPS PPP AR was mainly tested.According to the statistical results of five days,static PPP AR showed a small improvement since the float PPP already has satisfying accuracy and convergence time;while kinematic PPP AR demonstrated a significant improvement both in accuracy and convergence time.For instance,the horizontal error component is about 1cm compared with 2cm,and the vertical component is better than 4cm compared with 5cm,and the convergence time has been shortened at least 50%.Besides,we also used the observation data from MGEX(Muti-GNSSEXperiment)to test the PPP AR results of GPS and Galileo systems.The GPS results still show an improvement in convergence time,but the kinematic fixed solutions are not as stable as the related float solutions,which may be caused by the low accuracy of calculated satellite UPDs.Since the number of observed satellites of Galileo system is relatively small,the kinematic solutions are quite noisy.We only tested its static PPP AR.But the results were not satisfactory.We investigated the reasons and found lower accuracy of calculated narrow lane UPDs.Hence,it is necessary to improve the accuracy of estimated narrow lane UPD.Further more,PPP AR of combined satellite system also needs to be investigated.更多还原