【作者】 王硕；

【导师】 陈卫东；

【作者基本信息】 中国科学技术大学 ， 电磁场与微波技术， 2014， 博士

【摘要】 分布式无源雷达成像主要是指利用空间上充分展开的多发射机和多接收机同时进行目标观测的一种成像方式,着重强调的是多发射机和多接收机之间相对于目标的空间展开性。本文分别从正向和逆向两个角度对分布式无源雷达成像问题进行分析,通过正向分析,文中研究了基于非均匀傅立叶变换的成像方法。通过逆向分析,研究了基于正则化的成像方法。在结合目标场景的稀疏先验后,研究了基于压缩感知的成像方法,对该方法下的重构性能进行了分析与优化,并对成像方程的失配问题进行了研究。本文首先从信号和几何的角度建立了分布式无源雷达的成像模型,表明了空间谱域回波数据与目标散射系数之间的傅立叶变换关系。进而从空间谱的角度对分布式无源雷达的成像性能进行了分析,并给出了空间谱填充样式与收发布站之间的一般关系。依据空间谱域的非均匀填充特性,提出了基于非均匀傅立叶变换的分布式无源雷达成像方法,并通过仿真对成像的性能进行了分析。研究表明,该方法能够在空间谱填充处于相对均匀且致密的条件下获得良好的成像结果。为获得更为一般条件下的分布式无源雷达成像方法,本文从逆问题的角度对成像过程进行分析,表明成像逆问题本质上是一个病态问题,并对成像逆问题病态性的特点进行研究,给出了病态性程度与收发布站之间的一般规律。为减弱病态性的影响,提出了基于Tikhonov正则化的分布式无源雷达成像方法,为削弱重构矩阵中系统扰动的影响,又提出了基于截断总体最小二乘的分布式无源雷达成像方法,通过仿真对成像的性能进行了分析与比较。针对实际中可获得的发射机和接收机数目有限,相应的空间谱填充处于稀疏条件的问题,本文对基于压缩感知的分布式无源雷达成像问题进行了研究,将几种经典的重构算法应用于成像反演中,并对正交匹配追踪算法在噪声条件下的重构特性进行了深入研究,推导并分析了迭代参数的取值要求以及成功重构的概率。进而对分布式无源雷达稀疏成像中的布站优化问题进行了研究,并通过并仿真对优化后重构性能的提升进行了验证。最后,本文分别对分布式无源雷达稀疏成像中的两类重构矩阵失配问题进行研究。针对发射机和接收机位置误差导致的失配问题,通过建立含有位置误差的成像模型,分析了位置误差对成像过程的影响特点,提出了基于辅助散射点的位置误差估计与校正方法,针对成像场景网格划分导致的重构矩阵失配问题(即Off-Grid问题),提出了基于谱估计的直接求解散射点位置的方法,并通过仿真对上述方法进行了验证。更多还原

【Abstract】 Distributed passive radar imaging mainly refers to a kind of imaging method that can make use of transmitter and receiver for target observation at the same time. The transmitter and receives are distributed on spatial domain. It is emphasized the spatial extensibility of the transmitter and receives to the target. Distributed passive radar imaging problems were analysised from the angles of froward and reverse respectively. Through positive analysis, this paper studies the imaging method based on non-uniform Fourier transform. Through reverse analysis, the method of imaging based on regularization was studied. After combing the sparse prior of target, imaging method based on compressive sensing is studied, and the performance of the method are analyzed and optimized, and the mismatch problem of imaging equation was studied.First of all, distributed passive radar imaging model is established in this article from the perspective of the signal and the geometry. The imaging model shows the Fourier transform relationship between the target scattering coefficient and the echo data in spatial spectrum. The distributed passive radar imaging performance is analyzed in the perspective of spatial spectrum. The relationship between the spatial spectrum filling and the station layout is studied. Study and put forward the distributed passive radar imaging method based on the non-uniform Fourier transform, and analyzes the performance of the method through the simulation, the research shows that the method can get a good imaging results in the uniform and dense space spectrum filling condition.This article study the distributed passive radar imaging problem from the perspective of inverse problem in more general condition. This paper suggests that imaging inverse problem is essentially an ill-posed problem. Analyzed the imaging characteristics of inverse ill-posed problem, and presents a general law between the ill degree and the station layout. Study and put forward a distributed passive radar imaging based on Tikhonov regularization method to weaken the influence of the ill-posed problem. A distributed passive radar imaging method based on truncated total least squares is proposed in order to weaken the system disturbance in reconstruction matrix. The imaging performance is analyzed and compared through the simulation.The distributed passive radar imaging problem based on compressive sensing is studied for the station number is limited in the actual environment, and the corresponding spatial spectrum filling in a state of sparse. Several classical algorithm was applied to imaging inversion. The reconstruction characters of orthogonal matching pursuit algorithm under the condition of noise carried on the research. The iterative parameter values and the probability of successful reconstruction is derived. The station layout optimization problem is studied under the condition of sparse imaging. The optimization performance is analyzed and compared through the simulation.Finally, two class reconstruction matrix mismatch problem in the sparse distributed passive radar imaging problem is studied respectively. The influence of position error on imaging process is studied based on the imaging model containing position error. The position error estimation and correction method based on auxiliary scattering point is proposed for the mismatch problem caused by position error. This paper proposes a direct solving scattering point location method based on spectrum estimation technique for mismatch problem caused by Off-Grid, and is verified by simulation.更多还原