基于稀疏域高分辨表示的海杂波抑制及微动目标检测技术研究

The complex motion of a marine target will result in the non-linear property of Doppler with micromotion characteristics. However, the micromotion characteristics are extremely weak due to the strong sea clutter and their energy are overlapped in both time and frequency domain, which would increase the difficulty of radar detection. The research intends to improve the performances of detection and estimation for weak target in sea clutter. The micromotion characteristics of sea clutter and marine target in sparse domain are employed for sea clutter suppression and target detection. At first, based on the analysis of sparse property of radar returns, sea clutter suppression operator are designed via the high-resolution sparse representation technique, only using the data in the test unit. Then, the influence of non-stationary and homogeneous sea clutter on target detection can be reduced. Morphological component analysis and sparse time-frequency decomposition are used to distinguish sea clutter and micromotion target, and extract characteristics as well. Finally, the differences of sparse characteristics between sea clutter and micromotion target are employed to form test statistics for constant false alarm rate detection. The performances of the proposed algorithms are verified using real data. Compared with the traditional time-frequency analysis and transform domain methods, the proposed methods have obvious advantages of time-frequency resolution, anti-clutter, efficiency, and multicomponent signals analysis. The research will provide novel ways for improvement of detection ability for marine weak target.

海上目标复杂运动导致多普勒频率随时间非线性变化，具有微动特征，然而强海杂波的存在使得微动特征极其微弱，两者在时域和频域均有交叠，增加了雷达探测难度。本项目以改善海杂波背景中弱目标检测和参数估计性能为目标，利用海杂波和海面目标微动特征的稀疏性进行杂波抑制和目标检测。在分析海面回波信号稀疏性的基础上，采用高分辨率稀疏表示技术，仅利用待检测单元观测数据设计稀疏域海杂波抑制算子，有效降低非平稳非均匀海杂波对目标检测性能的影响，改善信杂比。采用形态成分分析和稀疏时频分解的方法区分海杂波和微动目标，提取微动特征。最后，利用海杂波与微动信号的稀疏特性差异，构造检测统计量，实现目标恒虚警检测，并基于实测数据对算法进行验证与性能分析。相比传统时频分析以及变化域处理方法，所提方法在时频分辨率、抗杂波、运算效率以及多分量信号分析等方面具有明显优势，为进一步提高雷达海上弱目标的探测能力提供新的思路。

海杂波背景下稳健可靠的微弱动目标检测始终是雷达信号处理领域的一个难题。本项目研究了稀疏域海杂波抑制和动目标检测方法，完成了项目计划书中预定的全部研究目标。针对不同的观测条件、雷达配置方式和目标运动特性建立微动信号模型，并从变换域和稀疏分解角度分析海杂波及海上微动目标回波的稀疏特性及影响因素；提出了基于短时分数阶傅里叶变换的海面目标微动特征检测和提取方法、短时稀疏时频分布（ST-STFD）框架；基于稀疏优化的短时稀疏分数阶表示域微动目标检测方法、基于稀疏傅立叶变换的快速算法以及稳健的稀疏分数阶变换方法，并用于微动目标检测。经实测雷达数据验证，所提方法具有抗杂波干扰能力强、时频分辨率好、运算效率高等特点。. 该项目共发表学术论文25篇，其中SCI收录2篇，EI收录16篇。申请国家发明专利6项，授权2项。项目负责人入选了中国科协“青年人才托举工程”，获得国际无线电联盟（URSI）青年科学家奖、中国专利优秀奖、中国产学研军民融合奖等科研奖励，组织承办了3次大型学术会议，在本领域会议作报告32次，研究成果受到了国内外同行的认可。