合成孔径雷达成像、干涉与定位精度的退化机理及补偿方法研究

System errors and motion errors as well as approximately processing may degrade the accuracy of SAR image, interferogram and three-dimensional location. Exploring a research on the mechanism of accuracy degradation and deriving algorithms to compensate these errors could improve image quality，interferogram accuracy and three-dimensional location accuracy during SAR data processing, and hence provide more credible data for the applications such as inversion, image translation and topographic mapping. Since its great value, this project will conducte a research on accuracy degradation mechanism and its compensation method of Synthetic Aperture Radar (SAR) imaging, interferometry and location. First, the research comprehensively analyzes the effect of inaccuracies such as motion error, system error and approximation error during the whole data processing flow including imaging, interferometry and location. Then it set up a mapping model between the sources of error and the accuracy of SAR data product to reveal the accuracy degradation mechanism, and provide a scientific basis for the design of compensation algorithm. Second, residual motion errors modeling, autofocus and dynamic sub-aperture division methods are employed for developing the compensation approaches of residual motion error. Third, the interferogram generation method based on modern spectrum estimation will be developed to improve interferogram accuracy. Finally, joint calibration methods based on united sensitivity matrix and optimization model will be proposed, and the weights will be introduced to these calibration methods. In conclusion, this project will help enhance the accuracy of SAR imaging, interferometry and location, and may contribute to the development of SAR data processing theory.

系统误差、运动误差和近似处理的存在，会导致SAR成像、干涉和定位的精度出现偏差。探索其精度的退化机理，并针对性地设计补偿方法，有助于提升SAR成像、干涉和定位的精度，进而为后续的反演、解译、制图等应用提供更加可靠的数据。鉴于其重要意义，本项目围绕SAR成像、干涉和定位精度的退化机理及补偿方法展开研究。首先，全面分析干涉SAR系统中各种误差源在整个数据处理流程中的作用机理，建立误差源与精度退化之间的映射模型，揭示精度退化的机理，为补偿算法的设计提供依据；其次，采用残差建模、自聚焦与动态子孔径划分相结合的方法，研究具备保相特性的残余运动误差补偿方法；再次，研究基于现代谱估计思想的干涉相位生成方法；最后，研究基于整体敏感度矩阵和最优化模型的区域网加权定标算法。总之，本项目的研究成果将提升SAR成像、干涉与定位的精度，同时有助于催生SAR数据处理的新思路、新方法，进而推动SAR数据处理理论的发展。

SAR 成像、干涉和定位精度的提升，可以为后续的反演、解译、制图等应用提供更加可靠的数据。鉴于其重要意义，本项目围绕SAR 成像、干涉和定位精度的退化机理及补偿方法展开研究，主要研究内容包括：（1）建立了IMU测量误差、系统延时误差、多普勒中心频率误差、参考DEM误差等因素与残余运动误差之间的函数关系，阐明了SAR残余运动误差的产生与影响机理；（2）建立了相控阵体制SAR系统误差的模型，并设计了误差补偿方法，该方法能够有效补偿SAR中央电子设备和相控阵天线的非理想特性所引入的幅频、相频误差；（3）从产生机理出发，建立了多路径误差的数学模型，并给出了误差补偿方案；（4）对SAR无源干扰、噪声压制干扰、民用射频干扰、有源相参干扰的作用机理进行了分析，并提出了针对性的干扰抑制方法；（5）基于现代谱估计思想中的AR模型，推导了干涉相位生成方法，该方法能够不进行精配准而直接估计SAR干涉相位；（6）提出了基于整体敏感度矩阵的联合定标算法和基于最优化模型的联合定标算法，提高了定标精度和自动化程度；（7）在SAR干涉制图中，提出了三维定位精度和后向散射精度的联合校准方法，保证了各像素点的几何位置精度和后向散射精度，保证了邻接场景之间的位置和灰度衔接性；（8）提出了InSAR多次观测数据的融合方法，通过融合减少了DEM的随机误差。总之，本项目的研究成果能够提升SAR 成像、干涉与定位的精度，同时有助于催生SAR 数据处理的新思路、新方法。