网络RTK快速定位算法研究与实现
文献类型：学位论文
| 作者 | 罗孝文 |
| 学位类别 | 博士 |
| 答辩日期 | 2008-12 |
| 授予单位 | 中国科学院测量与地球物理研究所 |
| 授予地点 | 武汉 |
| 导师 | 欧吉坤 |
| 关键词 | GPS 中长基线 网络RTK 相位整周跳变 相位模糊度 病态问题 正则化方法 选权拟合法 系统误差 三步分离法 虚拟参考站 流动站 快速定位 |
| 学位专业 | 大地测量学与测量工程 |
| 中文摘要 | 网络RTK是最近几年高精度GPS实时定位研究的一个热门话题，它在20世纪九十年代中叶被提出。网络RTK指的是在一定的区域内建立多个基准站(一般为三个或三个以上)，所建立的基准站对这个区域构成网状结构，并以一个或多个基准站为基准，计算和模拟各种空间相关的误差，并发布各种改正信息，对该地区的卫星定位用户所测值进行实时改正的一种测量方式，因此又称为多基准站的RTK。网络RTK包括数据传输系统和数据管理系统，需要管理实时产生的改正数，原始观测数据，每个观测站的多路径改正以及精密或预测的IGS轨道数据等等。在网络RTK的发展历程中，很多学者进行了多方面的研究。目前面临的问题主要是如何进行网络RTK数据的快速处理。本文在深入研究了网络RTK定位技术的理论基础以及各种误差的处理方法的基础上，对参考站和流动站的数据快速处理进行了以下有益的探索，并对一些关键问题提出了新的见解。本文的主要工作和成果包括以下几个方面的内容： (1)把选权拟合方法应用到网络RTK的数据处理中 在网络RTK参考站数据快速处理中，由于观测值比较少，在求解过程中就会出现法方程的病态而导致求解的不稳定。为了解决这个问题，引入了选权拟合方法，同时结合参考站坐标的先验信息，克服了法方程的病态性。选权拟合方法的引入为网络RTK的研究提供了一种新的研究思路。 (2)提出了一种网络RTK参考站间整周跳变实时探测的方法 分别以 ， 载波历元间的三差观测量经过宽巷( ),消电离层( )线性组合，结合中长基线参考站坐标已知的先验信息，利用选权拟合方法求解三差“模糊度”的浮点解，在此基础上用LAMBDA方法搜索出整周跳变(三差“整周模糊度”)，实现中长基线参考站整周跳变的实时探测，解决了网络RTK整周跳变实时探测的难题。 (3)对目前网络RTK参考站间的模糊度求解进行探讨，提出一种网络RTK参考站间模糊度快速求解的方法 现有的解法大多是采用卡尔曼滤波把模糊度和大气误差一起进行估算，这些方法存在解算复杂,所需要的时间较长的不足。针对这个问题， 以选权拟合方法作为求解模糊度的主要方法，此时求得的模糊度浮点解靠近其准确整数值，为搜索模糊度的整数解提供了方便，实现了模糊度的快速求解。 (4)在分析各种系统误差处理方法的基础上，提出了一种系统误差三步分离的方法 许多学者对系统误差的处理进行了研究，但是快速的系统误差求解，特别是在卫星升降时候系统误差的求解是网络RTK要解决的一个难题。本文将经过双差消电离层组合的码与相邻历元相位差值作为基本观测量，把码和相位的系统误差当作待估参数，利用选权拟合方法首先估算出码的系统误差，接着用去掉系统误差的码观测值和相位估算出相位的系统误差，随后利用估计的相位系统误差辅助模糊度的求解，在模糊度固定好以后，实现系统误差的精确求解。 (5)探讨了流动站模糊度的求解，提出了一种快速的模糊度求解方法 目前，如何快速求解流动站的模糊度仍是当前研究和需要解决的问题。问题本质是求解一个精度比较高的初始化坐标。本文提出了一种快速的模糊度求解方法，以 和 的组合为基本的观测值，通过卡尔曼滤波求解宽巷模糊度和协方差阵，再通过改进的LAMBDA方法进行模糊度的搜索，实现了宽巷模糊度的快速求解。利用宽巷模糊度求解出精度比较高的初始化坐标，再次进行线性化，调用选权拟合方法实现了载波L1模糊度的快速求解，实现了厘米级的快速定位。通过实验验证，本方法实现了模糊度的快速求解，而且有比较高的可靠性。 |
| 英文摘要 | One of hot themes of high precision GPS real time positioning, network RTK, originally brought forward in the mid-1990s and its idea is that several (three or more) reference stations are constructed, which form a network. Based on one or more reference stations, people can calculate and model the errors correlative with space, as well as pass various information and correct users’ observations within this area. So it also named multi- reference stations RTK. Network RTK is made up of two main parts: data transmissions system and data manage system. It manages the real time corrected number and raw observations, and multi-path corrections and precise or forecast IGS orbit information and so on. Many researchers study network RTK from various aspects. At present, the main problem lies in how to treat with the data come from network RTK quickly. On the base of lucubrated the theory about orientation technology of network RTK and disposal method of various errors, this paper searches after the quick processing that data from reference and rover station, and gives some new sights about several key problems. The following is the main work and achievements of this paper. (1) Apply the selecting weight fitting to data processing of network RTK Because of the less observation, the ill-conditioned of normal equation would be appeared to cause unsteadiness of solution. In order to solve this problem, we import the selecting weight fitting method to combine with propri coordinates of reference stations’. It can overcome the ill-conditioned of normal equations. The introducing of select weight fitting supply a new research thought for network RTK. (2) Advancing a new method to detect the cycle slip among network RTK reference stations Using three difference observations between epochs of the linearity combinations of wide lane (L5) and ionosphere-free (L3), and basing on the known reference stations’ coordinates of medium-long baseline, we estimate the floating solution by using selecting weight fitting method. And then use LAMBDA method to search the cycle slip (three differences “integer ambiguity”). That is, the puzzle about detecting the cycle slip of medium-long baselines’ reference stations in real time is resolved. (3) Discussing the estimation of ambiguity among network RTK reference stations and putting forward a fast algorithm Existing methods applying Kalman filter to estimate ambiguity and atmospheric errors together. The shortcomings of these methods are complex and needing longer time. We adopt the selecting weight fitting as main method to estimate the floating solution. This floating solution is approach to accurate integer ambiguity. So it is more convenient to search the integer ambiguity and to realize the fast estimation of ambiguity. (4) Advancing a new method to separate systematic errors in three steps Although many scholars study the estimation of systematic errors, it is still a difficult problem especially at the moment of satellites are rising or falling. In this paper, the code, coming from ionosphere-free combination by double difference, and phase difference between two adjacent epochs as basic observations, and the systematic errors of code and phase as unknown parameters, we still use selecting weight fitting to estimate codes’ systematic errors, and then use codes, which were separated from systematic errors, and phases to estimate the systematic errors of phase, which is then to assistant to estimate the ambiguity. This method can fasten the estimation of ambiguity and increase the success rate. After estimation of ambiguity, we can calculate systematic errors precisely. (5) Proposing a fast algorithm to estimate the ambiguity of rover station At present, how to estimate the ambiguity of rover station is still a pending subject. Its essence is to get more accurate initialization coordinates. In this paper, a fast algorithm would been introduced, it taking combination of L6 and L5 as basic observations, and using Kalman filter and improved LAMBDA method to search the wideline ambiguity. After that, more accurate initialization coordinates are gotten and linearization redone. L1 ambiguities are resolved by selecting weight fitting. Experimental studies verified this algorithm can estimate the ambiguity fast and reliably. |
| 公开日期 | 2013-01-16 |
| 源URL | [http://ir.whigg.ac.cn//handle/342008/3661]  |
| 专题 | 测量与地球物理研究所_学生论文_学位论文 |
| 推荐引用方式 GB/T 7714 | 罗孝文. 网络RTK快速定位算法研究与实现[D]. 武汉. 中国科学院测量与地球物理研究所. 2008. |
入库方式： OAI收割
来源：测量与地球物理研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。