PPP中基于CORS网的区域对流层延迟拟合模型

发布时间：2018-07-12 09:36

本文选题：对流层延迟改正 + 区域拟合模型　；参考：《长安大学》2013年硕士论文

【摘要】：精密单点定位技术（PPP）是利用GPS的后处理精密星历以及精密卫星钟差，以单台双频GPS接收机采集的相位数据作为主要观测值来进行单点定位的，其精度可达厘米级。随着精密单点定位技术的提高，PPP数据处理中的一些误差源受到了人们的重视，其中卫星钟差、电离层延迟和对流层延迟是GPS精密单点定位中三个关键误差源。因此获取高采样间隔的精密钟差产品、建立电离层二阶项改正模型、获取高精度的对流层延迟改正量是当下研究热点。本文介绍了卫星钟差和电离层延迟的最新研究进展，主要对PPP中对流层延迟改正的基本理论和实现方法进行了深入的学习和研究。对流层延迟是精密单点定位中一项重要的误差，它的估计精度直接影响最终的定位精度。在精密单点定位中，一般先对对流层延迟进行模型估计，再将剩余残差进行参数估计。随着GPS连续运行参考站（CORS）的建立，基于CORS网的对流层信息能够提供高精度的对流层延迟改正估计量，通过这些数据可以建立对流层延迟改正的区域拟合模型，然后对拟合模型进行内插获取流动站的对流层延迟改正信息。这种方法不但可以提高PPP的定位精度，，而且可以加快PPP的收敛速度。目前常用的区域拟合方法有一次曲线拟合、三参数平面拟合、二次曲面拟合、多面函数拟合等。然而上述拟合方法只考虑了拟合模型中因变量的误差，没有考虑自变量的误差。基于以上问题，本文提出了两种同时考虑因变量和自变量误差的拟合模型，分别为改进的三参数拟合模型和基于整体最小二乘法的一元线性拟合模型。通过观测数据建立的数学模型，由于实际问题的复杂性和观测数据不可避免的含有误差等原因，导致建立的模型往往不精确，因此需要对模型进行补偿。本文通过最小二乘配置方法对一次曲线拟合模型进行补偿，建立了基于最小二乘配置的一次曲线拟合模型，并通过自编程序对上述三种新模型进行了验证分析。
[Abstract]:Precision single Point Positioning (PPP) is based on the GPS post-processing precision ephemeris and the precise satellite clock difference. The phase data collected by a single dual-frequency GPS receiver are used as the main observation values for single point positioning. The precision of PPP can reach centimeter level. With the improvement of precision single point positioning technology, people pay more attention to some error sources in PPP data processing, among which satellite clock error, ionospheric delay and tropospheric delay are the three key error sources in GPS precise single point positioning. Therefore, it is a hot topic to obtain precise clock difference products with high sampling interval, to establish the ionospheric second-order correction model and to obtain high-precision tropospheric delay corrections. In this paper, the latest research progress of satellite clock difference and ionospheric delay is introduced, and the basic theory and implementation method of tropospheric delay correction in PPP are studied. Tropospheric delay is an important error in precise single point positioning, and its estimation accuracy directly affects the final positioning accuracy. In precise single point positioning, the model estimation of tropospheric delay is usually carried out first, and then the residual error is used to estimate the parameters. With the establishment of the GPS continuous operation reference station (cors), the tropospheric information based on the cors net can provide a high-precision tropospheric delay correction estimator. Through these data, a region-fitting model of tropospheric delay correction can be established. Then the fitting model is interpolated to obtain the tropospheric delay correction information of the mobile station. This method can not only improve the positioning accuracy of PPP, but also accelerate the convergence rate of PPP. The region fitting methods commonly used at present include the first curve fitting, the three parameter plane fitting, the Quadric surface fitting, the multi-plane function fitting and so on. However, the above fitting method only considers the errors of dependent variables in the fitting model, and does not consider the errors of independent variables. Based on the above problems, this paper presents two kinds of fitting models which take into account the errors of dependent variables and independent variables at the same time. They are the improved three-parameter fitting model and the univariate linear fitting model based on the global least square method. Because of the complexity of practical problems and the inevitable error of observation data, the mathematical model established by observation data is often inaccurate, so it is necessary to compensate the model. This paper compensates the first order curve fitting model by the least square collocation method, establishes the first order curve fitting model based on the least square collocation, and verifies and analyzes the three new models by self-programming.
【学位授予单位】：长安大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P228.4

【参考文献】