GPS interactive time series analysis software

Time series analysis is an important part of geodetic and geodynamic studies, especially when continuous GPS observations are used to explore areas with a low rate of deformation. In this domain, having precise and robust tools for processing and analyzing position time series is a prerequisite. To meet this requirement, a new software package called GPS Interactive Time Series Analysis was developed using the MATLAB language. Along with calculating basic statistics and quality parameters such as mean and variance, the software is capable of importing and visualizing different time series formats, determining and removing jumps and outliers, interpolating data, and producing numerical and publication quality graphical outputs. Furthermore, bivariate statistical analysis (such as correlation coefficients, curvilinear and nonlinear regression), residual analysis, and spectral analysis (such as auto-spectrum, Lomb–Scargle spectrum, evolutionary power spectrum, and wavelet power spectrum) form the main analysis features of the software.     

CATS: GPS coordinate time series analysis software

Over the last 10 years, several papers have established that daily estimates of GPS coordinates are temporally correlated and it is therefore incorrect to assume that the observations are independent when estimating parameters from them. A direct consequence of this assumption is the over-optimistic estimation of the parameter uncertainties. Perhaps the perceived computational burden or the lack of suitable software for time series analysis has resulted in many heuristic methods being proposed in the scientific literature for estimating these uncertainties. We present a standalone C program, CATS, developed to study and compare stochastic noise processes in continuous GPS coordinate time series and, as a consequence, assign realistic uncertainties to parameters derived from them. The name originally stood for Create and Analyze Time Series. Although the name has survived, the creation aspect of the software has, after several versions, been abandoned. The implementation of the method is briefly described to aid understanding and an example of typical input, usage, output and the available stochastic noise models are given.     

A MATLAB-based Kriged Kalman Filter software for interpolating missing data in GNSS coordinate time series

GNSS coordinate time series data for permanent reference stations often suffer from random, or even continuous, missing data. Missing data interpolation is necessary due to the fact that some data processing methods require evenly spaced data. Traditional missing data interpolation methods usually use single point time series, without considering spatial correlations between points. We present a MATLAB software for dynamic spatiotemporal interpolation of GNSS missing data based on the Kriged Kalman Filter model. With the graphical user interface, users can load source GNSS data, set parameters, view the interpolated series and save the final results. The SCIGN GPS data indicate that the software is an effective tool for GNSS coordinate time series missing data interpolation.     

最小二乘拟合GNSS位置时间序列分析

针对GNSS位置时间序列包含的线性趋势及其变化可能干扰后续分析并掩盖动力学因素信息的问题,该文使用最小二乘方法分析时间序列的线性趋势并减弱时间序列中阶跃的影响。分析了最小二乘方法用于GNSS位置时间序列分析的可行性及利用该方法分别获取趋势变化点前后的线性趋势,据此估计时间序列趋势变化的大小进而可以对时间序列进行修复。该方法用于GNSS位置时间序列的初步分析,可以方便有效地去除线性趋势变化对后续时间序列分析的影响,同时拟合结果本身也能反应出时间序列的变化特征。     

联合EMD-HD和小波分解的GNSS坐标时间序列降噪分析EI北大核心CSCD

针对经验模态分解(empirical mode decomposition,EMD)在GNSS坐标时间序列的降噪过程中存在筛选准则的选取和模态混叠效应等问题,本文引入Hausdorff距离(Hausdorff distance,HD)筛选准则并结合小波分解(wavelet decomposition,WD),提出EMD-HD&WD算法。通过对我国大陆构造环境监测网络149个GNSS测站的垂向坐标时间序列降噪处理,分别利用复合指标T值、测站的速度不确定度和闪烁噪声振幅验证算法的可靠性和普适性。结果显示:HD优于现有的筛选准则;EMD-HD&WD算法对测站的速度不确定度和闪烁噪声振幅的平均改正率均为88.4%。分析表明,本文算法能够有效识别和剔除噪声并且改善EMD的模态混叠效应,提高GNSS垂向坐标时间序列的模型精度。     

华北区域GNSS基准站高程时间序列的非线性变化分析

基于陆态网络全球卫星导航系统（GNSS）观测成果,采用功率谱分析法和最小二乘方法,以华北地区为例,研究了区域基准站高程时间序列的非线性变化特征,并分析了不同环境负载的影响.结果表明,GNSS基准站高程方向存在显著的周年和半年周期特征,且周年特征要显著于半周年特征.位于不同地区的基准站的振幅和相位存在差异,华北平原南部地区的周年振幅要大于北部地区,整体上华北地区周年变化在秋季时节振幅达到最大.不同环境负载效应对华北GNSS高程位移的影响不一致,利用三种环境负载修正GNSS序列后,水文负载的修正效果最好,非潮汐大气负载次之,非潮汐海洋负载修正结果不理想.      

Bayesian methods for outliers detection in GNSS time series

This article is concerned with the problem of detecting outliers in GNSS time series based on Bayesian statistical theory. Firstly, a new model is proposed to simultaneously detect different types of outliers based on the conception of introducing different types of classification variables corresponding to the different types of outliers; the problem of outlier detection is converted into the computation of the corresponding posterior probabilities, and the algorithm for computing the posterior probabilities based on standard Gibbs sampler is designed. Secondly, we analyze the reasons of masking and swamping about detecting patches of additive outliers intensively; an unmasking Bayesian method for detecting additive outlier patches is proposed based on an adaptive Gibbs sampler. Thirdly, the correctness of the theories and methods proposed above is illustrated by simulated data and then by analyzing real GNSS observations, such as cycle slips detection in carrier phase data. Examples illustrate that the Bayesian methods for outliers detection in GNSS time series proposed by this paper are not only capable of detecting isolated outliers but also capable of detecting additive outlier patches. Furthermore, it can be successfully used to process cycle slips in phase data, which solves the problem of small cycle slips.     

基于支持向量机的GNSS时间序列预测

在深度学习的理论框架下,针对预测全球卫星导航系统(GNSS)时间序列,传统的经验风险最小化预测模型误差大精度低,泛化性能差且对历史数据的经验依赖大的问题.提出一种采用结构风险最小化原则的基于支持向量机(SVM)的时间序列预测模型.通过和多层的BP神经网络预测模型预测效果比较,结果证明SVM预测模型拥有更好的时间序列预测效果.     

基于独立分量法的新疆GNSS时间序列共模误差分析CSCD

共模误差(CME)是区域连续全球卫星导航系统(GNSS)网中的主要误差来源之一.针对GNSS时间序列具有非高斯分布特征,基于二阶统计量的主成分分析(PCA)难以准确提取出CME分量问题,采用具有高阶统计量的独立分量分析(ICA)对CME进行提取.以2011—2018年新疆区域GNSS坐标时间序列为例,将PCA滤波效果进行对比验证,分析了CME对GNSS坐标时间序列的影响,并对CME序列进行周期分析.结果表明:前6个独立分量包含CME分量,这可能与卫星轨道、地表质量负荷和时钟误差有关,ICA滤波后东(N)、北(E)、天顶(U)三个方向的均方根(RMS)值分别降低31.83%、32.29%、35.49%,速度不确定度分别降低44.14%、38.49%、43.32%,各测站的周期项振幅较滤波前更一致,有效地剔除了CME,提高了坐标时间序列的精度.     

三种GNSS高程时序降噪方法的效果对比分析

为了探究经验模态分解(EMD)、整体经验模态分解(EEMD)和小波降噪三种方法的降噪性能,以中国区6个国际GNSS服务(IGS)站高程分量的5?a、10?a和20?a时序数据为例,对它们的降噪结果进行比较分析.?首先利用线性拟合分离趋势项,并采用3σ准则剔除异常值,得到满足符合降噪要求的样本序列;然后分别用这三种方法分...     

GNSS高程时间序列周期项的经验模态分解提取

针对传统方法——常振幅常相位的周年+半周年谐波模型法不能准确提取GNSS高程时间序列中周期项问题,该文以中国区域10个IGS基准站在ITRF2008框架下2005—2015年高程时间序列为例,采用经验模态分解(EMD)提取各测站高程时间序列的周期项。对传统方法和EMD两种方法提取的周期项做Lomb_Scargle谱分析,用功率谱图分析了这两种方法提取序列周期项的能力。实验结果表明,EMD方法较传统方法更能准确、自适应地提取GNSS高程时间序列的周期项。     

利用CEEMD分析中国沿海GNSS站高程时序变化

利用完备经验模态分解方法（CEEMD）对我国沿海地区6个GNSS基准站（2010—2018）的高程时序数据进行了处理分析。结果表明：CEEMD在高程时间序列分析中具有一定的优越性,可准确分解出各GNSS站高程时序中存在的周、月、季节、年等变化周期项,其中周年运动是主要贡献项,各站高程时间序列的短周期变化与潮汐变化周期具有密切关联性;沿海GNSS站的地面沉降既具有区域的一致性,又存在区域间差异性,其中D区DBJO、DZJJ站呈现先下降后上升的趋势,N区NZUH、NWZU站呈下降趋势,B区的BZMW呈上升趋势,而同海区的BLHT站则呈显著的下降趋势。     

京津地区GNSS坐标时间序列噪声特性研究

针对地表质量负荷对京津地区GNSS坐标时间序列噪声特性的影响,选取中国大陆构造环境监测网络8个GNSS基准站2012—2014年的坐标时间序列,利用CATS软件计算大气压、非潮汐海洋、积雪和土壤湿度等质量负载改正前后GNSS坐标时间序列的谱指数、最优噪声模型、速度的变化。发现地表质量负载对GNSS坐标时间序列的噪声特性产生了明显影响。结果显示,京津地区GNSS坐标序列包含白噪声和有色噪声,且最优噪声模型具有多样性。扣除质量负载后N、U分量的噪声模型变化明显,主要表现为FN+WN和PL+WN,而N、E分量的谱指数分别趋近于FN和WN。质量负载改正后基准站U方向的线性速度变化较大,且北京地区变化量大于天津地区。研究结果为提高GNSS数据解算精度、精细分析地壳形变提供参考。     

联合LMD与EMD的GNSS站坐标时间序列去噪方法

针对GNSS站坐标时间序列信噪不易分离的问题,在传统EMD去噪方法的基础上,本文提出了一种联合LMD与EMD的坐标时间序列去噪方法。该方法首先采用LMD分解原始坐标时间序列,基于连续均方误差（CMSE）原则分离高频噪声与低频信号,保持低频分量不变;然后对高频分量进行EMD去噪;最后以2次分解所得低频信号之和作为去噪后时间序列。以仿真数据与8个GNSS基准站实测数据进行试验,通过多种评价指标进行精度评估。结果表明,与传统EMD方法相比,联合LMD与EMD的方法能够更加精确地去除坐标时间序列中的噪声。     

GNSS模糊度整数估计方法图形可视化软件设计与应用分析

模糊度快速准确估计是全球卫星导航系统(GNSS)高精度定位的关键,整数取整、序贯取整和整数最小二乘估计是模糊度常用的三类整数估计方法.尽管从程序上较易实现三类估计方法,但是如何根据模糊度浮点解和精度构建整数估值的几何图形却缺乏较多的研究,不利于我们对整数估计过程的直观认知.因此,本文从理论上分别给出三类估计方法的一般形式,然后基于MATLAB GUI设计了一套三类估计方法二维几何图形构建的可视化分析软件,其功能包括三类估计方法的归整域构建、映射图构建和蒙特卡洛模拟及成功率计算.实验测试结果表明,本文设计的软件能够从几何图形角度较直观地表达出三类整数估计过程及其解算性能.      

L1 regularization for detecting offsets and trend change points in GNSS time series

Offset and trend change point detection are major problems for GNSS time series preprocessing. Without accurate detection of change points and offsets, signals estimated from GNSS time series are prone to be biased. To solve this problem, we introduced an extensive L1 regularization model, which can estimate piecewise trends, level shifts and seasonal signals simultaneously from raw GNSS time series. It thus can be used to detect trend change points and discontinuities successfully in GNSS time series. Furthermore, a new Python tool has been incorporated into our previous TSAnalyzer software to realize the benefits our L1 regularization model and some examples are listed to show its usage.     

含缺值GNSS基准站坐标序列的非插值小波分析与信号提取

受多种因素影响,GNSS基准站坐标序列通常都含有缺值,传统小波分析需要对缺值数据进行内插或补零处理。本文基于小波系数与时间序列观测数据的重构关系,提出了一种非插值的二进小波变换的最小范数解法,导出了相应的计算式,并严格证明了传统的补零处理算法与本文的最小范数解法等价。最后利用中国地壳运动观测网络一期27个基准站实测数据以及模拟数据进行了验证分析。结果表明,本文的非插值算法与插值算法提取的信号差异较小,27个基准站坐标序列的平均残差中误差仅相差2.01%(North),0.54%(East)和1.26%(Up),两种算法提取的信号之差与信号平均方差比仅相差1.16%(North),0.54%(East)和1.62%(Up)。     

Correlation acceleration in GNSS software receivers using a CUDA-enabled GPU

GPS Solutions - The correlation process in a GNSS receiver tracking module can be computationally prohibitive if it is executed on a central processing unit (CPU) using single-instruction...     

LMD和小波阈值的GNSS坐标时间序列降噪应用

局部均值分解方法降噪过于粗糙,将认定为噪声的乘积函数(PF)分量直接剔除,导致有用信息丢失.为了有效提取GNSS站坐标时间序列的有用信息,该文提出一种局部均值分解和小波阈值相结合的降噪方法.通过局部均值分解将坐标时间序列分解为一系列PF分量和余项,依据消除趋势波动分析方法计算各PF分量的Hurst指数,利用小波阈值提取H≤1的PF分量中的有用信息,将提取出的信息与剩余PF分量叠加重构获得最终降噪的坐标时间序列.通过对5个测站的坐标时间序列进行实验,结果表明局部均值分解和小波阈值相结合的方法能够有效提取噪声分量中的有用信息,信噪比提高了27.8％,从而验证了该方法的有效性.