利用基于抗差垂直向方差分量估计的GPS-InSAR数据融合方法反演三维形变场

GPS-InSAR数据融合解算三维形变场模型易受观测值粗差影响,且基于方差分量估计的定权方法不具备抵御粗差能力,计算效率低下。鉴于此,本文提出了一种基于抗差垂直向方差分量估计的GPS-InSAR融合解算模型,利用方差分量估计方法及抗差估计理论,通过对观测值最优化分类并进行选权迭代,精确分配权重,进而有效计算三维形变场。试验结果表明,该方法能有效抵御观测值粗差不利影响,提高三维形变场反演精度,提升逐点式计算的三维形变场效率。

GPS-InSAR data fusion method with robust vertical variance component estimation for 3D deformation field

It is easily affected for the three-dimensional deformation field model of GPS-InSAR data fusion by the gross error of the observation value. The weighting method based on the Helmert variance component estimation cannot be used to resist gross errors and the calculation efficiency is low. In view of this, a method based on the GPS-InSAR fusion solution model for robust vertical variance component estimation is proposed. The variance component estimation method and the robust estimation theory are used to calculate the 3D deformation field effectively by optimizing the classification of the observed values and carrying out the weight selection iteration to allocate the weight accurately. Experimental results show that it is effectively resisted the influence of gross errors in observations by this way. The accuracy of 3D deformation field inversion and the efficiency of 3D deformation field calculated point by point is effectively improved.