单历元GPS变形监测数据处理理论与相关方法研究

【作者】 王齐林；

【导师】 朱建军； 邹峥嵘；

【作者基本信息】 中南大学 ， 大地测量学与测量工程， 2003， 硕士

【摘要】 由于GPS载波相位相对定位的高精度，使它成为变形监测中极其重要的监测手段，特别是在桥梁、大坝以及其它建筑物的变形方面有广阔的应用；随着GPS硬件和软件技术的成熟，实时自动化的高精度变形系统将是今后的发展趋势，这种系统能够实现数据采集自动化、内外业处理一体化。在这样的系统中，数据的实时处理是系统建设中的关键问题之一。 论文着重于GPS数据的单历元解算方法研究，以及监测点变形量的分析和预测，此外，还探讨了这种方法在变形监测系统数据处理中的应用情况以及数据的相关性对变形精度的影响。首先，论文系统地分析了单历元算法的解算过程，研究了在不同条件下固定模糊度和提取变形量的方法，分析了约束条件下的单历元解算方法和解算模型的解算过程。试算结果表明，这种算法和模型可以很容易的求得监测点的变形量。接着，本文阐述了对变形值进行分析和预测的各种方法，提出了用整体方差法和假设检验的方法来对变形量进行分析，并用卡尔曼滤波和回归分析方法对变形值进行预测，试算结果表明这些方法是可行的。 数据之间的相关性将影响观测成果，在GPS变形监测系统中，这种影响不可忽视。论文分析了观测值之间的数学相关性，给出了数学相关下的权阵，阐述了观测数据之间的物理相关性，并根据基于时间序列的残差值，求出了时间相关性的函数式，相关时间为200s以上。论文还对空间相关性进行了分析，阐述了多路径效应的影响。根据变形监测的特点以及单历元解算方法，论文最后初步设计了一个GPS变形监测系统。这个系统可以采用单历元算法和变形量的分析方法，同时在数据处理过程中考虑到数据的相关性。更多还原

【Abstract】 Because of its high precision of GPS carrier-phase relative position, GPS becomes an important method of construction deformation monitoring, especially widely applied in bridge or dam deformation and so on. Along with the progress of technique of GPS hardware and soft, real time automatic monitoring system with high precision will be the trend, it can realize the automation of collecting data and integration of work inside and outside. During the constructing of those systems, real time data processing is one of critical problems.The paper emphasizes on the method of GPS single epoch algorithm and the influence on result of deformation survey because of correlation among observations, the paper researchs the application of the algorithm of data processing in deformation monitoring system, and the influence on precision of deformation monitoring. Firstly, The paper analyses procedure of data processing of single epoch algorithm, and researchs the method about how to fix ambiguity and pick-up deformation value, it also analyses algorithm with constraints and model of single epoch data processing. According to the result of example, this algorithm with constaints and model are effective and practicable to get deformation value of the monitoring points. Secondly, the paper elaborates many ways to analyze and forecast deformation value, and proposes to use total variance methods and hypothesis and test to analyze, and use regressing and Kalman filter to forecast deformation information. It is confirmed by the example.Correlation will affect the result, and it should not be neglected in defonnation monitoring system. Lastly, The paper analyses mathematical correlation between data, and gives weight matrix of mathematical correlation, and elaborates physical correlation, gives function of time correlation according to residual errors based on time serial, the correlation time is above 200 seconds, and analyses influence of time correlation. This paper also analyses spatial correlation, and elaborates the effect of multi-path. According to characteristic of deformation system and single epoch algorithm, the paper designs a system primaryly, it uses single epoch algorithm and data analyzing methods, andconsidering data correlation at the same time.更多还原