利用高频GPS单站数据解算同震地表位移关键算法研究

发布时间：2018-01-16 16:18

本文关键词：利用高频GPS单站数据解算同震地表位移关键算法研究　出处：《西南交通大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着高频GPS(Global Positioning System)接收机的逐渐普及,以及采用高频GPS技术可直接获取测站位移形变等优势,使得高频GPS技术在地震学领域有了广阔的发展空间。本文根据地震发生时间很短暂,而短时间内对流层延迟等误差强相关性和没有周跳时模糊度固定不变的特点,通过采用单站时间基线模型获得地震期间测站瞬时地表位移。本文以时间基线模型和GPS单站星间单差消电离层模型为基础,得到单站星间单差时间基线观测模型,并对其随机模型进行了推导分析。基于单站星间单差时间基线观测模型,得出影响模型解算精度的主要是卫星钟差和周跳这两个因素。为进行卫星钟差改正,对分段线性插值方法、滑动拉格朗日插值方法和三次样条插值方法进行了分析,通过实验比较了三种插值方法的插值效果,实验表明:分段线性插值方法的精度最高。本文对电离层残差法及Wubbena-Melbourne组合法这两种常用的周跳探测与修复方法进行简要分析,通过实验验证了两种方法的探测效果及局限性;基于巴尔达数据探测法和 LAMBDA(Least-square Ambiguity Decorrelation Adjustment Method)方法,本文提出了基于状态估计的序贯最小二乘周跳探测与修复方法,并对该方法的各种影响因素进行了研究与分析。采用1Hz静态数据和1Hz地震数据进行周跳模拟实验,实验结果表明:该方法可以成功的探测与修复静态数据和地震数据中1周的小周跳。完成卫星钟差改正和周跳探测与修复后,本文采用5Hz的高频静态数据、EI-Mayor-Cucapah地震期间1Hz高频数据、芦山地震期间50Hz高频数据进行解算。实验结果表明:对于静态数据,采用本文方法解算得到的N、E、U三个方向的均方根误差分别为3.1mm,3.3mm,8.5mm,且位移结果分布图呈现正态分布的特性;对于采用本文方法所得EI Mayor-Cucapah地震期间的解算结果,相对于SOPAC(Scripps Orbit and Permanent Array Center)的双差解,本文方法不受地震期间参考站位移的影响,所得动态位移结果精度更高、更准确;芦山地震期间的GPS数据解算测站最大振幅结果与前人的研究成果相差仅为2mm。三个实验的结果进一步验证了本文单站星间单差时间基线模型的正确性和可靠性。
[Abstract]:With the increasing popularity of high frequency GPS(Global Positioning system receivers, and the advantages of using high frequency GPS technology to obtain the displacement deformation of measuring stations directly. The high frequency GPS technology has a broad development space in the field of seismology. According to the earthquake occurrence time is very short. In a short period of time, the tropospheric delay has strong correlation and the ambiguity is invariant when there is no cycle slip. The instantaneous surface displacement of the station during the earthquake is obtained by using the single station time baseline model. This paper is based on the time baseline model and the single difference ionospheric model of GPS single station. The single difference time baseline observation model is obtained, and the stochastic model is deduced and analyzed. The model is based on the single difference time baseline observation model. In order to correct the satellite clock error, the piecewise linear interpolation method is obtained. The sliding Lagrange interpolation method and the cubic spline interpolation method are analyzed, and the interpolation effects of the three interpolation methods are compared by experiments. The experiment shows that:. The precision of piecewise linear interpolation is the highest. In this paper, two commonly used cycle slip detection and repair methods, the ionospheric residual method and the Wubbena-Melbourne combination method, are briefly analyzed. The effectiveness and limitation of the two methods are verified by experiments. Based on Valda data detection and Lambda (. The Least-square Ambiguity Decorrelation Adjustment method. In this paper, a sequential least square cycle slip detection and restoration method based on state estimation is proposed. The influence factors of this method are studied and analyzed. The cycle slip simulation experiment is carried out by using 1Hz static data and 1Hz seismic data. The experimental results show that this method can successfully detect and repair the small cycle slip in the static data and seismic data. After the satellite clock error correction and cycle slip detection and repair are completed, the high frequency static data of 5 Hz are used in this paper. 1 Hz high frequency data during EI-Mayor-Cucapah earthquake and 50 Hz high frequency data during Lushan earthquake are calculated. The experimental results show that: for static data. The root mean square error of the three directions is 3.1 mm / 3.3mm / 8.5mm, respectively, and the distribution of the displacement results shows the characteristics of normal distribution. The calculated results of ei Mayor-Cucapah earthquake are obtained by using this method. Compared with SOPAC(Scripps Orbit and Permanent Array center. The method in this paper is not affected by the displacement of reference station during earthquake, and the result of dynamic displacement is more accurate and accurate. The difference between the maximum amplitude of the GPS data and the previous research results is only 2 mm. The results of three experiments further verify the correctness of the single difference time baseline model between single station and satellite in this paper. Reliability.
【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P228.4;P315.7

【参考文献】