| 引用本文: | 刘 慧， 郭馨宇， 郭子夜， 程碧辉. 基于交替方向乘子法的Capon层析SAR成像方法[J]. 雷达科学与技术, 2023, 21(3): 303-313.[点击复制] |
| LIU Hui, GUO Xinyu, GUO Ziye, CHENG Bihui. Capon Tomographic SAR Imaging Method Based on Alternating Direction Method of Multipliers[J]. Radar Science and Technology, 2023, 21(3): 303-313.[点击复制] |
|
| 摘要: |
| 合成孔径雷达（Sythetic Aperture Radar， SAR）层析成像（TomoSAR）是一种多基线干涉测量技术，可沿垂直于视线（Perpendicular to the Line?Of?Sight, PLOS）方向估计功率谱图（Power Spectrum Pattern， PSP）即后向散射系数，从而实现三维成像。本文提出一种改进的波束形成优化算法，在双约束鲁棒Capon波束形成算法（Doubly Constrained Robust Capon Beamforming， DCRCB）的基础上，结合L1范数的约束函数，构建交替方向乘子法（Alternating Direction Method of Multipliers， ADMM）的代价函数，将DCRCB恢复的后向散射系数进行进一步稀疏优化，实现层析SAR的三维成像。ADMM算法以增广拉格朗日算法为基础，将较为复杂的全局求解问题转换为两个或多个更易求解的简单局部子问题。ADMM算法在迭代中，各子问题可分别完成稀疏重构和降噪运算，被分离的局部子问题代数式都较为简单，均能较容易地求出确定的解，且不必对其进行收敛运算与约束操作。因此，ADMM算法具有重建精度高的优势。本文采用2021年中国科学院空天信息创新研究院发布的山西运城地区的8通道机载阵列干涉SAR数据进行了实验验证，实验结果验证了算法的有效性。 |
| 关键词: 合成孔径雷达层析成像 波束形成 交替方向乘子法 稀疏优化 |
| DOI：DOI:10.3969/j.issn.1672-2337.2023.03.009 |
| 分类号:TN957.52 |
| 基金项目:国家自然科学基金青年科学基金项目（No.61501019） |
|
| Capon Tomographic SAR Imaging Method Based on Alternating Direction Method of Multipliers |
|
LIU Hui, GUO Xinyu, GUO Ziye, CHENG Bihui
|
|
School of Electrical and Information Engineering，Beijing University of Civil Engineering and Architecture, Beijing 102616, China
|
| Abstract: |
| Synthetic aperture radar （SAR） tomography （TomoSAR） is a multi?baseline interferometric technology, which can estimate the power spectrum pattern （PSP） or backscattering coefficient along the direction perpendicular to the line?of?sight （PLOS） to achieve three?dimensional imaging. This article proposes an improved beamforming optimization algorithm. Based on the doubly constrained robust Capon beamforming （DCRCB） algorithm and the constraint function of L1 norm, the cost function of the alternating direction method of multipliers （ADMM） is constructed to further optimize the backscatter coefficient recovered by DCRCB and then realize three?dimensional imaging of tomographic SAR. The ADMM algorithm is based on the augmented Lagrangian algorithm, which transforms the more complex global solution problem into two or more local sub?problems that are easier to solve. In the iteration of ADMM algorithm, each sub?problem can complete sparse reconstruction and noise reduction operation respectively. The algebra formula of the separated local sub?problem is relatively simple, and it is easy to obtain the determined solution without convergence operation and constraint operation. Therefore, ADMM algorithm has the advantage of high reconstruction accuracy. The capabilities of the proposed method are demonstrated by means of the 8?channel airborne array interferometric SAR data released by Aerospace Information Research Institute, Chinese Academy of Sciences acquired by the uninhabited aerial vehicle over the urban area of Yuncheng of Shanxi Province in 2021. The experimental results verify the effectiveness of the algorithm. |
| Key words: synthetic aperture radar tomography（TomoSAR） beamforming alternating direction multiplier method（ADMM） sparse optimization |
