| 题名 | 加速GPS精密单点定位收敛的方法研究 |
| 作者 | 郝明 |
| 学位类别 | 硕士 |
| 答辩日期 | 2007-06 |
| 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 | 精密单点定位 加速收敛的方法 选权拟合法 自适应选权滤波 |
| 学位专业 | 大地测量学与测量工程 |
| 中文摘要 | GPS精密单点定位（Precise Point Positioning，简称PPP）是利用GPS精密卫星星历及精密卫星钟差，以单台双频GPS接收机采集的非差相位数据作为主要观测值，进行单点定位计算，其精度可达分米甚至厘米级水平。这种技术除了可以进行精密定位，还可以进行低轨卫星精密定轨，对流层和电离层的监测，应用前景广阔，是目前GPS界研究的热点之一。 本文首先总结和归纳了精密单点定位的国内外发展现状及研究意义；然后系统地介绍了精密单点定位的理论基础，包括IGS产品，GPS观测值及其线性组合，精密单点定位中的主要误差源和关键技术；详细的阐述了精密单点定位的数学模型，包括函数模型和随机模型。 针对目前精密单点定位收敛时间比较长这一问题，本文探讨了一种加速精密单点定位收敛的新方法。由于初始状态参数及其方差-协方差阵选取不准，以及在卡尔曼滤波过程中有些状态参数的预报可能发生异常，所以PPP一般需较多历元的观测数据才能达到分米或厘米级水平，收敛时间比较长。本文在对PPP算法深入研究的基础上提出了一种新方法加速PPP收敛。新方法采取了两步措施： ①将选权拟合法应用到初始状态参数及其方差-协方差选取上。选权拟合法的基本思想是如果事先对于部分参数有较可靠的先验信息，通过对部分参数附加适当约束，则会改善解的结果。通过分析PPP中状态参数的特点，恰当的选取约束条件，可以得到比较准确的卡尔曼滤波初始值及方差-协方差矩阵。 ②将自适应选权滤波应用到卡尔曼滤波参数估计中。当观测值出现粗差或实际状态与模型假设不一致，卡尔曼滤波结果会出现大的偏差甚至严重失真，还会导致滤波的发散。本文将自适应选权滤波应用于卡尔曼滤波参数估计中，根据状态参数的特性，将不正常的个体分辨出来，然后对这些不正常个体适当降权，进而分别形成相应的等价权，取代原有状态参数的权，在不改变原卡尔曼滤波算式的形式的基础上得到合理的结果。 最后，通过算例表明采用新方法比改进前定位精度有明显提高，在比较短的时间内定位精度可达分米甚至厘米级，验证了本文方法的可行性和有效性 |
| 英文摘要 | Precise Point Positioning (PPP) uses un-differenced dual-frequency pseudorange and carrier phase observations along with IGS precise orbit and satellite clocks, for stand-alone point positioning (static or kinematic) with decimeter to centimeter level accuracy. Besides precise positioning, PPP has been applied to a variety of geophysical phenomena, including precise orbit determination of low earth orbiter (LEO), ionosphere and troposphere sounding, etc. This paper describes the investigated meaning and development of PPP in home and abroad; summarizes the fundamental knowledge of PPP systematically; discusses the mathematical model of PPP detailedly, including functional model and stochastic model. Because of the convergence time of present PPP is long, this paper introduces a new method of improving convergence of PPP. While determining initial values of state parameters and corresponding variance-covariance matrices, they may not be accurate. And in the process of Kalman filtering, some of the state parameters may forecast abnormally. So PPP may be need many observations to achieve decimeter or centimeter positioning accuracy, the time of convergence is long. Basing on the investigation of PPP algorithm, a new method is put forward to improve convergence: ①The selecting weight fitting method is applied to determine the initial values of state parameters and corresponding variance-covariance matrices. The basic idea of selecting weight fitting method is that if a priori information of some parameters is known, we could give appropriate constraints as additional to improve the numerical results. By analyzing the characteristic of the state parameters, appropriate constraints could get more accurate initial values of state parameters and corresponding variance-covariance matrices. ②The adaptive filtering by selection of the parameter weights is used in Kalman filtering. While gross errors appearing in observations or real state disaccording with supposed model, there will be great mistakes in Kalman filtering results, and they would cause disconvergence in filter. According to the characteristic of state parameters, the abnormal individual is distinguished and deweighted suitably. Then corresponding equivalent weight is formed, replacing the weight of previous parameters. Basing on the unchangeable situation of Kalman filtering, the adaptive filtering by selection of the parameter weights could achieve reasonable consequences. At last, the numerical experiments demonstrate that the accuracy of PPP could achieve decimeter to centimeter level in a short time, and the feasibility and validity of the method presented in this paper can be ensured. |
| 公开日期 | 2013-01-22 |
| 内容类型 | 学位论文 |
| 源URL | [http://ir.whigg.ac.cn//handle/342008/3698]  |
| 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 | 郝明. 加速GPS精密单点定位收敛的方法研究[D]. 武汉. 中国科学院测量与地球物理研究所. 2007. |
| 个性服务 |
| 查看访问统计 |
| 相关权益政策 |
| 暂无数据 |
| 收藏/分享 |
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。
修改评论