基于低秩与稀疏矩阵分离的视频合成孔径雷达 GMTI信号处理技术研究

GMTI signal processing technology of ViSAR system is an advanced subject in the field of the microwave remote sensing. It possesses the ability of high resolution and high frame rates to realize moving target surveillance, which has broad application prospects in military and civilian aspects. To overcome the inefficiency and insufficiency problem suffered from the traditional GMTI algorithm and inter-frame difference algorithm, the theory of low-rank and sparse matrix separation is introduced for the GMTI signal processing of ViSAR system in this project. The main contents of the project include four aspects. Firstly, system parameters of ViSAR system are optimized. Secondly, echo signal models of ViSAR system are constructed, and the radar echoes in different conditions are simulated; the low-rank property of ground clutter and sparse property of moving targets are revealed. Thirdly, the detailed signal processing chain is designed based on low-rank and sparse matrix separation theory, including moving target extraction algorithm, parameter estimation algorithm and moving target location algorithm. Finally, the performances of the designed algorithms are investigated; the influence of different parameters to the final precision of moving target location is analyzed, and the effectiveness of the proposed methods is confirmed. Through the studies in the project, a fast and efficient ViSAR moving target extraction method is expected to form, which lays the theoretical and technical foundations for the application of ViSAR-GMTI technology in the civil and military areas. Consequently, this project has the important theoretical significance and application prospect.

视频合成孔径雷达GMTI信号处理技术是微波遥感领域的前沿研究课题，具有高分辨率、高帧率的运动目标监视能力，在军用和民用领域均具有重要的应用价值。本课题拟将低秩与稀疏矩阵分离理论和视频合成孔径雷达GMTI信号处理技术相结合，解决传统GMTI算法以及帧间差分算法在处理视频合成孔径雷达GMTI数据时效率低、信息挖掘不充分的问题。课题的研究内容主要包括：优化视频合成孔径雷达系统参数；构建视频合成孔径雷达回波模型，生成各种情况下的场景回波,揭示杂波的低秩性与运动目标的稀疏性；设计基于低秩与稀疏矩阵分离的运动目标提取算法，以及基于提取结果的参数估计与定位算法；研究运动目标提取算法的性能，分析各种参数对定位精度的影响，验证整个信号处理流程的有效性；形成一套完整的视频合成孔径雷达运动目标信息快速、高效获取方法，奠定视频合成孔径雷达GMTI技术在实际应用的理论和技术基础。课题具有重要的理论意义和应用前景。

视频合成孔径雷达GMTI信号处理技术是微波遥感领域的前沿研究课题，具有高分辨率、高帧率的运动目标监视能力，在军用和民用领域均具有重要的应用价值。本课题将低秩与稀疏矩阵分离理论和视频合成孔径雷达GMTI信号处理技术相结合，解决了传统GMTI算法以及帧间差分算法在处理视频合成孔径雷达GMTI数据时效率低、信息挖掘不充分的问题。课题的研究内容主要包括：优化视频合成孔径雷达系统参数；构建视频合成孔径雷达回波模型，生成各种情况下的场景回波,揭示杂波的低秩性与运动目标的稀疏性；设计基于低秩与稀疏矩阵分离的运动目标提取算法，以及基于提取结果的参数估计与定位算法；研究运动目标提取算法的性能，分析各种参数对定位精度的影响，验证整个信号处理流程的有效性；形成一套完整的视频合成孔径雷达运动目标信息快速、高效获取方法，奠定视频合成孔径雷达GMTI技术在实际应用的理论和技术基础。在本课题的支持下，共发表 EI 收录论文 4 篇，中文重要核心收录论文2篇，授权发明专利1项，申请发明专利3项。此外，课题负责人还获得2018年国防科技进步一等奖1项，排名第五。