GPS动态变形监测数据处理模型研究

发布时间：2018-04-25 03:18

本文选题：GPS + 变形监测　；参考：《中国矿业大学》2014年硕士论文

【摘要】：随着越来越多高耸建筑物的兴建，其结构健康状况受到人们普遍的关注，及时掌握建筑在台风、地震等外界因素作用下的形变信息和受损程度，可以为建筑物的安全运营提供保障。和传统监测手段相比，GPS在变形监测方面具有高频、实时和不受外界环境干扰等特性，使得GPS观测技术成为变形监测一种重要的新方法。本文围绕GPS动态变形监测展开研究，主要内容如下：（1）针对复杂环境伪距噪声较大的情况，提出了一种周跳探测及修复的组合算法，并实验论证了算法的有效性和实用性。利用开源代码RTKLib的相对定位模块，设计短基线解算程序，并与Track软件进行对比实验，验证了程序解算的结果具有较高的精度。（2）详细介绍了对流层延迟改正的基本方法，从温度、相对湿度、大气压三个方面，模拟其变化和高差对对流层延迟的共同影响。在Kalman滤波中附加天顶对流层延迟参数，利用随机过程法进行估计，设计无高差和高差较大的两组基线进行对比实验，分析改正效果。（3）首先分别从伪距多路径和测站坐标序列的角度出发，分析测站连续两天多路径效应的相关性，论证多路径效应的周日重复性。设计自适应滤波器分离连续两天的坐标信号，其相关部分即为多路径误差，首先通过模拟实验验证自适应滤波器分离信号的有效性，，然后对实测数据进行实验，分析所设计的自适应滤波器在多路径误差处理中的效果。对于单一信号或者相关性不好的输入信号，自适应滤波不再适用，提出采用小波分析提取多路径效应，通过振动台实验，分析原始信号中的测量噪声、振动信号、周期多路径效应以及低频多路径效应和温度综合影响等。（4）以香港某高楼在台风荷载作用下实测的数据为例，分析以上理论方法在动态变形监测中的适用性及有效性。提取高楼结构振动信号时，先利用主成分分析PCA，联合区域内同步观测数据，分离共模误差并消除站间相关残余误差，然后利用小波分析方法提取高楼结构振动信号，并从频率域分析提取振动信号的正确性。
[Abstract]:With the construction of more and more high-rise buildings, people pay more and more attention to the health of their structures. They can grasp the deformation information and damage degree of the buildings under the influence of typhoon, earthquake and other external factors in time. Can provide the safeguard for the safe operation of the building. Compared with the traditional monitoring methods, GPS has the characteristics of high frequency, real time and no external environment interference, which makes the GPS observation technology become an important new method of deformation monitoring. This paper focuses on the dynamic deformation monitoring of GPS. The main contents are as follows: 1) in view of the large pseudo-range noise in complex environments, a combined algorithm for cycle slip detection and restoration is proposed, and the validity and practicability of the algorithm are demonstrated experimentally. Using the relative positioning module of open source code RTKLib, a short baseline program is designed and compared with Track software. The results show that the program has a high accuracy. 2) the basic method of tropospheric delay correction is introduced in detail. The influence of its variation and height difference on tropospheric delay is simulated from three aspects: temperature, relative humidity and atmospheric pressure. The zenith tropospheric delay parameters are added to the Kalman filter, and the stochastic process method is used to estimate them. Two groups of baselines with no height difference and great height difference are designed to carry out comparative experiments, and the correction results are analyzed. Firstly, from the point of view of pseudo-range multipath and station coordinate sequence, the correlation of multipath effect for two consecutive days is analyzed, and the daily repeatability of multipath effect is proved. The adaptive filter is designed to separate the coordinate signal for two consecutive days. The related part is the multipath error. First, the validity of the adaptive filter separation signal is verified by the simulation experiment, and then the experimental results of the measured data are carried out. The effect of the designed adaptive filter in multipath error processing is analyzed. For the single signal or the input signal with poor correlation, adaptive filtering is no longer applicable. Wavelet analysis is used to extract the multipath effect, and the measurement noise and vibration signal in the original signal are analyzed by shaking table experiment. Periodic multipath effect, low frequency multipath effect and temperature comprehensive effect, etc. Taking the measured data of a tall building in Hong Kong under typhoon load as an example, the applicability and effectiveness of the above theoretical methods in dynamic deformation monitoring are analyzed. When extracting the vibration signal of tall building structure, the principal component analysis (PCA) is used to combine the synchronous observation data in the region to separate the common mode error and eliminate the residual error between the stations. Then the vibration signal of the tall building structure is extracted by wavelet analysis. The correctness of vibration signal is analyzed and extracted from frequency domain.
【学位授予单位】：中国矿业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：P228.4;TU196.1

【参考文献】