通过精化分析GPS坐标序列来提升与拓展GPS的地球物理应用已成为国际上的前沿热点研究，而如何准确有效地量化分离GPS坐标序列中的线性与非线性信息是该研究的核心。本项目拟重点突破GPS坐标序列中非线性瞬态变形信息提取方法，以及GPS测站精确线性速度及其不确定度的求解方法。具体地：首先，构建一个顾及地球物理机制的GPS坐标时序仿真平台，以此作为后续算法研究的数据源；然后，发展基于盲源分离技术的瞬态弱变形信息提取新算法；进而，在修正瞬态变形非线性信息的基础上，提出能抵抗各种残余污染信息的速度及其不确定度的非参数估计方法；最后，通过典型应用来研究认识相关地球物理现象。本项目的实施不仅可为科学解释我国板块构造运动、地壳变形，精化地球参考框架提供理论技术支持，还可为分析其它大地测量时序数据提供新思路。

Through refining analysis the GPS coordinate time series to improve and expand the GPS geophysical application has become an international frontier and hotspot research. And how to accurately and effectively quantify the linear and nonlinear information in the GPS coordinate time series is the key issue of the research. This project aims to break through the method of extracting the nonlinear transient deformation in the GPS coordinate time series and the method of solving the accurate linear velocity and its uncertainty of the GPS station. Specifically included: Firstly, a simulation platform will be build to prepare research data for new algorithms; Secondly, new algorithms to detect weak transient deformation by blind source separation (BBS) method will be developed; Thirdly, after correcting the nonlinear transient deformation, a nonparametric and robust estimation method is proposed to obtain the GPS velocity and its uncertainty.; Finally, we try to interpret some geophysical phenomena by typical applications. The implementation and accomplishment of our project will not only provide theoretical support for scientific explanation of plate tectonics, crustal deformation, refining terrestrial reference frame, but also provide a novel ideas for the analysis of other geodetic time series data.