再论拟准检定法的原理、实施和应用

综合阐述了拟准检定法的原理和特点、研究思路。拟准检定法的关键是如何正确选择拟准观测，文章介绍了初选的复选拟准观测的实施要点。列举了拟准检定法在图相关情况下的相差检测，形变分析中的异常探测以及GPS相位观测的周跳检测和修复等方面的应用例子。     

选权拟合与拟准检定

本文指出了拟准检定法和LEGE法的本质区别,分析了选权拟合法和拟准检定法各自的数学模型,发现拟准检定法是选权拟合法的一种特殊形式,于是,以选权拟合法原理为理论基础,推导了拟准检定真误差估值的计算公式。与以往的计算公式不同,其中的权使用了最小二乘改正数的权。但是由于可靠性矩阵是秩亏矩阵,改正数的权无法计算,所以直接采用观测值的权,此时,得到的公式才与以往的公式相同。     

粗差的拟准检定法(QUAD法)

以往对付粗差的各种方法都是以余差（残差）为研究对象的。本文提出一种全新的研究思路和方法,从真误差入手。真误差与观测值有确定的解析关系。但关系式的系数阵是秩亏的。本文借鉴周江文的“拟稳平差”思想,提出“拟准观测”的概念,附加“拟准观测的真误差范数极小”的条件,解决了关于真误差的秩亏方程组求解定解的问题,推导了粗差的拟准检定法算式。该方法依据真误差估值的分布特征,检测粗差准确可靠。不仅能有效地同时定位     

拟准观测的选取和真误差估值的“分群”现象

以往检验粗差有两类不同的方法,一类是假设检验,另一类是抗差估计。它们虽然有显著不同,但都是以最小二乘残差及其函数为研究对象的。文献［8,9］提出了以真误差为研究对象的拟准检定法。该方法通过附加拟准观测真误差的估值极小的条件,直接求解关于真误差的秩亏方程,然后依据真误差估值的分布特征判别和定位粗差。 　拟准检定法的关键是如何正确的选择拟准观测。本文总结了实践中选择拟准观测的原则和方法。选择拟准观测可采用初选与复选两步来实施。初选时将观测值分为4类：“0”类是按一定规则经过初步判断怀疑含粗差可能性很大的,不宜选为拟准检测;“1”类是依内部可靠性指标衡量,所处位置结构差的观测;“2”类与“0”相反,是依一定指标判断含粗差可能性较小的观测;“3”类是除上述三种特殊情况以外,余下的观测。这四类观测中,“2”类以及“3”类中指标值ui相对较小的部分观测可初选为拟准观测。复选是在初选后计算出真误差估值的基础上,选取真误差估值绝对值较小的为拟准观测。本文指出拟准观测的初选可以不唯一。 　利用许多统计学者曾多次讨论过的一个算例［6,11］,详细介绍了选择拟准观测的过程,并将有关结果与以往的结果进行比较,用拟准检定法一方面排除了4个异常观测的影响,另一方面充分利用了其余观测的信息,结果可能更合理一些。 　拟准观测选择适当时,真误差估值有明显的“分群”现象。结合算例,本文用图表的形式形象地分析了这种有趣的现象。在表中,异常的和正常的真误差估值的量值有显著的分界带,含粗差的或异常观测的真误差明显大于正常的;在图中,异常值“浮”在分布图的上部,而正常值则“沉”在下方。依此特性可直观地判定粗差的位置。     

用拟准检定法探测和修复GPS数据中的粗差和周跳

用粗差的拟准检定法来探测和修复GPS相位观测中的粗差和周跳，概述了用拟准检定法探测和修复周跳的数学模型，并给出了具体实施步骤，用两个算例验证本文方法的可行性和有效性。     

三种粗差检测方法的比较及分析

通过对数据探测法、拟准检定法和多维粗差同时定位定值法进行比较研究,对算例进行数值分析,比较这三种方法的优缺点。     

基于拟准检定法的多波束测深粗差探测

针对最小二乘趋势面滤波法在多波束测深粗差探测中存在的不足,本文提出将拟准检定法应用于多波束粗差探测中,进而将拟准检定法探测粗差与最小二乘趋势面拟合有机结合起来。通过构造不同阶次的趋势面,并对实测数据进行分析,结果表明:拟准检定法剔除粗差后海底地形表面较为平滑,微小地形得以保存,跳点和孤立点得以剔除。     

均值漂移模型在地壳运动异常测站剔除中的应用

地壳运动及形变分析时,一般认为块体运动以刚性运动为主,整体具有长期稳定性。但受块体局部构造运动和观测粗差影响,少数测站会出现异常运动,只有剔除这些异常测站才能获得对研究区域地壳的长期运动趋势及运动规律的合理分析结果。长期以来,在进行测站筛选时,大多采用将残差标准差的验后估计与残差相比是否超过某一限值来判别异常测站,做法不具备检测粗差的能力,对异常测站剔除并不充分。本文提出利用均值漂移模型对异常测站进行剔除,并与拟准检定法及传统基于残差标准差的简单LS估计法进行对比;对中国大陆构造环境监测网络在环渤海区域的近几期GPS观测的速度场数据进行拟合分析。结果表明,通过均值漂移量求解的迭代过程,不仅能起到抗差测站筛选的作用,而且与拟准检定法相比,还能避免拟准观测选择不准确造成的影响,从而获得更高速度场拟合精度。     

测量数据的质量控制理论探讨

从可靠性和粗差检测两方面分析了质控制理论研究的进展；介绍了新的统一的可靠性指标，它完全由系统的结构决定，不包含受主观因素影响的概率因子（非中心参数λ0)。详细介绍了拟准检定法的原理和实施办法，特别是初选拟准观测的一些技巧，通过一个算例比较了拟准检定法与假设检验的三种形式的计算结果，后者存在部分粗差不能发现的问题，而前者能发现全部粗差，效果较好。     

利用L1范数和中位数选取拟准观测值

粗差探测拟准检定法的核心是拟准观测的选取。提出了L₁范数和中位数相结合的方法选取拟准观测值,并设计了相应的准则。首先利用L₁范数方法得到稳健的残差,将其中残差接近于零时对应的观测值直接确定为拟准观测值,然后将余下残差形成新的残差向量,并计算其绝对值的中位数,拟准观测值即为那些余下残差绝对值小于中位数所对应的观测值。GPS网平差和GPS单点定位计算结果表明本文提出的选取拟准观测值的方法有效可行。     

Ability to detect and locate gross errors on DEM matching algorithm

Abstract

Digital elevation model (DEM) matching techniques have been extended to DEM deformation detection by substituting a robust estimator for the least squares estimator, in which terrain changes are treated as gross errors. However, all existing methods only emphasise their deformation detecting ability, and neglect another important aspect: only when the gross error can be detected and located, can this system be useful. This paper employs the gross error judgement matrix as a tool to make an in-depth analysis of this problem. The theoretical analyses and experimental results show that observations in the DEM matching algorithm in real applications have the ability to detect and locate gross errors. Therefore, treating the terrain changes as gross errors is theoretically feasible, allowing real DEM deformations to be detected by employing a surface matching technique.     

论真误差拟准解的基本特性

本文将回答这样一个问题：为什么拟准检定法能准确地定位粗差并估计出粗差大小？首先从理论上论证真误差的估值的重要特性：真误差拟准解的大小完全由它对应的观测值中的粗差决定,因此它能直接反映粗差的位置和大小。本文还用算例验证了这一结论。     

On separability for deformations and gross errors

According to the testing theory with two alternative hypotheses, the theory of detectability for a deformation monitoring network has been extended to the theory of separability for deformations and gross errors. With this theory it can be evaluated whether deformations and gross errors can be statistically distinguished each other, which is very important for deformation analysis. General formulas are established and the separability between deformations and gross errors for some typical deformation models is investigated. An example of separating gross errors from deformations is given.     

粗差拟准检定的实施方案设计

拟准检定法是根据真误差与观测值有确定的关系式RΔ=-RL,通过直接解算真误差估值Δ来检测粗差。由于真误差能直接反映粗差的大小，所以这种方法可靠性强。本文主要阐述该方法的实施方案设计思想及其步骤和选择“拟准观测”的一些技巧。     

Method for determining weight matrix for resolving three-dimensional surface deformation using Multi-LOS D-InSAR technology

Three-dimensional surface deformation is resolved using deformations obtained from multiple line of sight differential interferometry synthetic aperture radar (multi-LOS D-InSAR). However, the random model is susceptible to both accidental errors (such as errors due to noises) and system errors (such as orbital errors and topographic errors). To solve this problem, we present a weight-determining method based on high-precision global positioning system (GPS) observations. Specifically, the precision of multi-LOS D-InSAR deformation observations is evaluated based on high-precision GPS deformation observations from the same period. Then, by considering the influence of system errors and accidental errors on the deformation of a pixel observed by multi-LOS D-InSAR, a combined fitting model associated with the positions and elevations of a pixel is built. Finally, the precision of deformations from multi-LOS D-InSAR is evaluated pixel-by-pixel by using the built model. As a result, the weights of observations in the model for resolving three-dimensional surface deformation using multi-LOS D-InSAR can be determined. We verify the feasibility and precision of the weight-determining method through a simulation experiment and a real data experiment, respectively. The verification results show that the proposed weight-determining method based on high-precision GPS observations provides higher reliability for the resolved three-dimensional surface deformation.     

变形监测网的可区分性分析

本文首先借助于 Gauss-Markov 模型下两个多维备选假设检验的理论,将变形监测网单个备选假设下的灵敏度,在顾及模型误差的情况下,扩展成监测网的可区分性理论,提出了监测网的可区分性和区分可靠性概念;其次,对于监测网在三种典型变形模型(单点移动,错动和均匀应变)下变形与粗差的可区分性进行了讨论,获得了一些有益的结论。     

基于等价方差-协方差的粗差探测敏感度分析及实现

选权迭代法在面对独立观测量中的粗差时能够表现出良好的探测效果,但由于其只是一种基于独立观测值的稳健估计法,没有考虑到观测值之间的相关性[1]。而现有的等价权函数虽然都满足稳健估计的要求,但由于所构造的等价权阵不对称,使得最后平差结果严重偏离实际情况。本文介绍在传统稳健估计的基础上,在定权时充分考虑相关观测值之间相关性的不变性,构造对称的方差—协方差阵不断扩大,并通过VB进行编程及实例分析,发现该方法对粗差的敏感度非常强,探测精度很高。     

Gross Error Compensation for Gravity Field Analysis Based on Kinematic Orbit Data

This paper aims at a comparative study of several measures to compensate for gross errors in kinematic orbit data. It starts with a simulation study on the influence of a single outlier in the orbit data on the gravity field solution. It is shown that even a single outlier can degrade the resulting gravity field solution considerably. To compensate for outliers, two different strategies are investigated: wavelet filters, which detect and eliminate gross errors, and robust estimators, which due to an iterative downweighting gradually ignore those observations that lead to large residuals. Both methods are applied in the scope of the analysis of a 2-year kinematic CHAMP (challenging minisatellite payload) orbit data set. In various real data studies, robust estimators outperform wavelet filters in terms of resolution of the derived gravity field solution. This superior performance is at the cost of computational load, as robust estimators are implemented iteratively and require the solution of large sets of linear equations several times.