基于稀疏重构的SAR动目标检测技术

发布时间：2018-07-03 11:20

本文选题：合成孔径雷达 + 运动目标检测　；参考：《西安电子科技大学》2015年硕士论文

【摘要】：运动目标的检测和参数估计是雷达信号处理的重点方向。基于多通道合成孔径雷达(SAR)系统的运动目标检测方法能利用空域自由度实现有效的杂波抑制,完成地面慢速目标的检测(GMTI)。然而随着通道数目的增加,数据量急剧增大,造成雷达系统数据处理负担变大。压缩感知(CS)理论可以在欠采样条件下实现信号的无失真重构,并且运动目标信号在空间上通常具有稀疏性,因此,研究稀疏信号体制下的SAR动目标检测技术对提升雷达系统空间监视能力具有重要意义。本文主要研究基于稀疏重构的多通道SAR系统的动目标检测方法,主要内容如下:稀疏重构算法一般存在计算量过大的问题,难以直接应用于实际系统。本文借鉴零空间调整(NST)思想,提出一种快速的双通道SAR动目标检测方法。该方法联合双通道数据进行杂波抑制,利用NST算法实现动目标检测,提高了目标检测的精度和稳健性,并减小了算法的计算复杂度。动目标位置需标定在高分辨SAR图像上,尽管运动目标在空间上满足稀疏性,场景是非稀疏的,因此不能同时实现高分辨率SAR成像与动目标检测定位。针对该问题,首先对满采样通道采用交替方向法(ADM)重构场景和目标的SAR图像,利用所成SAR图像对稀疏采样通道进行杂波抑制,最后基于NST技术实现运动目标检测。仿真结果说明本算法具有良好的动目标检测性能。针对各通道稀疏重构时重组误差的不一致性,给出一种适用于稀疏采样模型的多通道动目标检测方法,大大提高了算法对重组误差的稳健性;在此基础上,提出一种稀疏采样多通道SAR的运动目标径向速度估计方法,该方法首先对各通道进行稀疏重构得到SAR图像,再通过对双通道数据进行ATI幅相联合检测实现运动目标定位,最终通过阵列DOA方法实现目标的高精度径向速度估计。
[Abstract]:Detection and parameter estimation of moving targets are the key points of radar signal processing. The moving target detection method based on multi-channel synthetic aperture radar (SAR) system can effectively suppress clutter by using spatial freedom and complete ground slow target detection (GMTI). However, with the increase of the number of channels, the amount of data increases rapidly, which makes the data processing burden of radar system become larger. Compression sensing (CS) theory can realize the distortion free reconstruction of the signal under the condition of under-sampling, and the moving target signal is usually sparse in space, so, It is very important to study the SAR moving target detection technology in sparse signal system to improve the space surveillance capability of radar system. In this paper, the moving target detection method of multi-channel SAR system based on sparse reconstruction is studied. The main contents are as follows: the sparse reconstruction algorithm is difficult to be directly applied to practical systems because of its large computational complexity. Based on the idea of null space adjustment (NST), a fast dual channel SAR moving target detection method is proposed in this paper. This method combines dual-channel data for clutter suppression and uses NST algorithm to realize moving target detection. It improves the accuracy and robustness of target detection and reduces the computational complexity of the algorithm. Moving targets need to be calibrated on high resolution SAR images. Although moving targets satisfy sparsity in space, the scene is non-sparse, so high resolution SAR imaging and moving target detection and localization cannot be realized at the same time. To solve this problem, alternate direction method (ADM) is used to reconstruct the scene and target SAR images, and the sparse sampling channel is suppressed by the SAR images. Finally, the moving target detection is realized based on NST technology. Simulation results show that the algorithm has good performance in moving target detection. Aiming at the inconsistency of recombination error in sparse reconstruction of every channel, a multi-channel moving target detection method suitable for sparse sampling model is presented, which greatly improves the robustness of the algorithm to the recombination error. A method for estimating the radial velocity of moving targets based on sparse sampling multi-channel SAR is proposed. Firstly, the SAR images are obtained by sparse reconstruction of each channel, and then the moving target location is realized by ATI amplitude-phase joint detection of two-channel data. Finally, high precision radial velocity estimation is realized by array DOA method.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN957.51

【参考文献】