基于CRInSAR和kalman滤波的监测地表三维形变的研究  

The study of estimating three-dimensional surface deformation using CRInSAR and Kalman filtering

作　　者：胡俊[1,3] 丁晓利[2] 朱建军[1] 李志伟[1,2,3] 张长书[1] 王兴旺[1] 

机构地区：[1]中南大学测绘与国土信息工程系,长沙410083 [2]香港理工大学土地测量与地理资讯学系,中国香港 [3]北京大学空间信息集成与3S工程应用北京市重点实验室,北京100871

出　　处：《测绘科学》2009年第2期50-52,共3页Science of Surveying and Mapping

基　　金：国家863计划项目(No.2006AA12Z156);国家自然科学基金(No.40404001);国家自然科学基金(No.40774003);空间信息集成与3S工程应用北京市重点实验室(北京大学)开放基金(项目编号:SIIBKL07-01-0)

摘　　要：CRInSAR技术克服了传统DInSAR的一些不足,成为近年来的研究热点,但单一的CRInSAR技术只能监测LOS的一维形变。而kalman滤波采用状态空间的概念,可以用来估计平稳或非平稳的多维信号随机过程,已经被广泛地应用于动态数据处理之中。因此本文以CRInSAR模型为基础,将不同时间跨度的干涉数据作为动态数据,构建相应的观测方程和状态方程,并以经典的kalman滤波估计角反射器的三维形变量和三维形变速度。实践证明,该方法是合理可行的。Because of conquering some deficiency of traditional DInSAR, CRInSAR has become an investigative hotspot recently. However, single CRInSAR just can survey one dimension deformation in the line of sight of radar. Kalman filtering, which introduces the conception of state space, could be used to estimate multidimensional signal random process in calm or un-calm state. And kalman filtering has been widely applied in dynamic data processing. In this paper, we used interferometric data in different time span as dynamic data and constructed corresponding observational equation and state equation based on CRInSAR model. Besides, we used classical kalman filtering to estimate 3D deformation and 3D deformation velocity of comer refiecter. It＇ s proved that this method is feasible by an experiment in the area of Hongkong

关 键 词：CRInSAR KALMAN滤波 动态 观测方程 状态方程 三维 

分 类 号：P228[天文地球—大地测量学与测量工程]