| 主成分分析模型及在变形监测中的应用 |
| |
| 引用本文： | 朱敏茹.主成分分析模型及在变形监测中的应用[J].北京测绘,2020(3):427-431. |
| |
| 作者姓名： | 朱敏茹 |
| |
| 作者单位： | 广州市城市规划勘测设计研究院 |
| |
| 摘 要： | 随机误差和多径效应作为GPS变形监测中的主要误差源,严重影响着GPS测量精度。针对这一问题,本文将主成分分析(Principal Component Analysis,PCA)模型引入GPS变形监测领域,首先利用传统PCA方法将测量数据转换至特征空间,通过剔除小特征值对应的特征向量实现对高斯分布随机噪声的抑制,然后将多径噪声作为色噪声进行分析,提出一种广义PCA方法利用多径噪声的时间相关性对其进行滤除,基于实际工程测试数据的实验结果表明,相对于传统的小波噪声抑制方法,所提方法可以获得更好的噪声抑制性能。
|
| 关 键 词： | 随机误差 多径效应 主成分分析 广义主成分分析 |
Principal Component Analysis Model and Application in Deformation Monitoring |
| |
| ZHU Minru.Principal Component Analysis Model and Application in Deformation Monitoring[J].Beijing Surveying and Mapping,2020(3):427-431. |
| |
| Authors: | ZHU Minru |
| |
| Institution: | (Guangzhou Urban Planning and Design Survey Research Institute, Guangzhou Guangdong 510000, China) |
| |
| Abstract: | As the main error sources of GPS deformation monitoring,random error and multipath effect seriously restrict the accuracy of GPS measurement.In order to solve this problem,this article will PCA(Principal Component Analysis,PCA)field model in GPS deformation detection,first of all,using the traditional PCA method will measure data conversion value feature space,by eliminating small eigenvalue corresponding eigenvector of Gaussian random noise suppression,and then the multipath noise as colored noise,a generalized PCA method is put forward on the use of multipath noise time correlation filter,based on the actual engineering test data of the experiment results show that compared with traditional wavelet noise suppression method,The proposed method can obtain better noise suppression performance. |
| |
| Keywords: | random errors multipath effect principal component analysis generalized principal component analysis |
| 本文献已被 维普 等数据库收录！ |
|
