国防科学技术大学硕士学位论文宽带雷达目标极化特征提取与识别研究姓名：邱伟申请学位级别：硕士专业：信息与通信工程指导教师：赵宏钟2010-11国防科学技术大学研究生院硕士学位论文 第 i 页 摘 要 雷达目标识别是现代雷达对目标实现定位、测量及辨识所亟需解决的重要研究课题，但由于问题本身的复杂性，至今仍无满意的解决方法。高分辨和全极化的结合被认为是雷达目标识别最具前景的研究方向之一，同时高分辨全极化雷达也成为现代雷达发展的重要方向。本文针对高分辨全极化雷达体制，深入开展了宽带雷达目标极化特征提取与识别的研究工作，主要内容包括基于极化目标分解的宽带雷达目标特征提取与识别、全极化散射中心参数估计、基于全极化散射中心分布的宽带雷达目标特征提取与识别等，具体安排如下： 第一章阐述了宽带雷达目标极化特征提取及识别研究的意义及现状，并给出本文主要研究工作简介。 第二章介绍了雷达目标极化散射特性的表征方式，提出一种结合极化目标分解和散射中心估计的雷达目标极化散射特性分析方法，暗室实测数据分析结果表明该方法能有效揭示目标的散射机理。 第三章在分析目标全极化高分辨距离像特性以及提取极化散射矩阵的基础上，首先利用 Krogager 分解、Cameron 分解和 Cloude 分解从雷达目标全极化高分辨距离像中提取三种极化特征，并设计了一种基于极化特征的目标识别流程，最后通过两组舰船目标电磁软件计算数据的识别实验检验极化特征的识别性能。与基于单极化高分辨距离像特征的识别性能相比较表明引入全极化信息能有效提高目标识别性能。 第四章首先根据不同极化通道下目标同一散射中心位置一致性的特点，研究了基于酉 ESPRIT 的全极化散射中心参数估计问题。 相对于常规的分极化通道散射中心单独估计方法，该方法只需一次运算就可估计出所有参数，估计的精度较高，而且避免了传统方法后续的散射中心筛选、配对等工作，极大地降低了运算量，提高了数据处理的效率。其次，鉴于舰船目标散射中心分布的稀疏性差异，提出了一种基于散射中心分布稀疏度特征的舰船目标识别方法，仿真实验结果表明，该特征具有良好的识别性能，验证了该方法的合理有效。 第五章对本文内容进行系统总结，并对下一步的工作进行了展望。 关键词：特征提取，目标识别，高分辨全极化雷达，高分辨距离像，极化目标分解，散射中心，稀疏性，极化散射特性 国防科学技术大学研究生院硕士学位论文 第 ii 页 ABSTRACT Radar target recognition is an important problem needed to be solved urgently for modern radar to implement target position, measurement and recognition. However, because of the complexity of the problem itself, there is still no satisfying solution. The combination of high resolution and full polarization is considered as the one of the most promising approach to solve the problem, and thus high-resolution fully polarimetric radar is also becoming an important direction of modern radar development. In this dissertation, wideband radar target polarimetric features extraction and recognition for the high-resolution fully polarimetric radar system is discussed in depth, mainly including fully polarimetric scattering center parameter estimation, wideband radar target polarimetric feature extraction and recognition based on the polarimetric target decomposition and distribution of scattering centers. The main work in this dissertation is organized as follows: The first chapter introduces the significance and status of wideband radar target polarimetric feature extraction and recognition, and gives the work profile of this paper. The second chapter describes the characterization polarimetric scattering characteristics of radar target ， and a method combinating polarimetric target decomposition and scattering centers estimation is proposed to analyze polarimetric scattering characteristics of radar target. This method is validated by real data measured in anechoic chamber. The third chapter firstly introduces the characteristics of fully high range resolution profile(HRRP) and polarimetric target decomposition theory. Subsequently, it applies the Krogager decomposition, Cameron decomposition and Cloude decomposition to the fully HRRP of radar target to extract three types of polarimetric features, based on which a target recognition procedure is designed; Finally, two simulation experiments on data of two groups of ships calculated through the electromagnetic software are carried out to test the effectiveness of these polarimetric features. Compared to the recognition performance of the feature extracted from single polarization HRRP, the method in this paper effectively improves the recognition performance by employing fully polarimetric information. The fourth chapter firstly works on fully polarimetric scattering center parameter estimation using Unitary ESPRIT according to the location consistency of scattering centers in different polarization channel. Unlike the conventional scattering center estimation under single polarization channel, the proposed method can not only estimate all parameters in one operation with high estimation accuracy, but also avoid such problems as screening and matching of scattering centers, which greatly reduces the computation burden and improves the processing efficiency. Secondly, for ship targets, 国防科学技术大学研究生院硕士学位论文 第 iii 页 given the sparseness differences of scattering centers distribution, a target recognition method based on sparse degree of scattering centers is put forward. In fact, it is observed from simulation experiment results that the feature of sparse degree can provide good recognition performance, which proves that the method is reasonable and effective. The last chapter presents a conclusion of this dissertation and a description of the future work. Key Words：feature extraction, target recognition, high-resolution fully polarimetric radar, high range resolution profile(HRRP), polarimetric target decomposition, scattering center, sparseness, polarimetric scattering characteristic国防科学技术大学研究生院硕士学位论文 第 III 页 表 目 录 表 2.1 几种典型目标的极化散射特性表征. 14 表 2.2 两极化基下极化散射矩阵. 18 表 2.3 0方位角下部分散射中心 Krogager 分解结果. 28 表 2.4 45方位角下部分散射中心 Krogager 分解结果.