光学区雷达目标结构成像的理论及其在雷达目标识别中的应用

【作者】 姜卫东；

【导师】 郭桂蓉； 庄钊文；

【作者基本信息】 中国人民解放军国防科学技术大学 ， 信息与通信工程， 2000， 博士

【摘要】 本文以宽带雷达体制为研究背景，在复杂背景下，主要研究了雷达背景杂波的仿真、雷达目标一维、二维结构成像的机理和方法，还研究了基于目标结构成像的识别方法等问题。 在序论中，首先简要回顾了目标电磁散射特性，以及雷达目标识别领域的研究方法和前人所作的工作，扼要总结了现代谱估计中谐波估计理论的现状和发展，介绍了本文研究的主要内容。 第二章以高分辨雷达杂波仿真为研究对象，以零记忆非线性变换(ZMNL)为研究手段，推演了几种典型杂波在ZMNL变换前后输入输出间非线性变换关系，提出了求解输入输出非线性关系式的快速数值求解方法(“误差控制”法)。回顾了产生均匀分布、高斯分布随机序列的数学原理，提出了基于AR模型的相关高斯杂波仿真方法。最后给出了多种杂波的仿真结果。 第三章以宽带阶梯变频雷达为研究背景，在白噪声和杂波条件下，研究了提取目标强散射中心的参数估计方法。首先从目标多散射中心理论出发，推演了描述目标散射特性的频域散射数学模型。在白噪声背景下，深入研究了几种典型的散射中心估计方法。在杂波背景下，研究了对杂波种类不敏感的散射中心估计方法。基于目标一维结构信息，提出了全极化条件下目标一维距离像的匹配识别方法；以目标散射中心为特征，提出了基于散射中心位置的相关匹配识别方法。最后用实验验证了个各算法的性能。 第四章在宽带体制下，推导了目标多散射中心的二维频域散射数学模型，分析了目标二维结构成像的两维分辨力和测量数据的坐标变换。在白噪声背景下，针对目标二维散射模型自相关函数“边缘解耦”特性，提出了估计目标二维散射中心的MP2D算法。在杂波背景下，研究了提取目标二维散射中心的RELAX算法。最后，用实验验证了两种算法的性能。 第五章总结了本文的研究工作，指出了需要进一步解决的问题。更多还原

【Abstract】 In this paper, under the wideband radar background, the simulation method of radar clutter, the principle and estimation method of one-dimension and two-dimension physical structure features of radar target, some classification method using physical structure features of target, etc. are studied.In chapter 1, the fields of electromagnetic scattering characteristics of target, the method and previous researched result in the radar target recognition domain, the present situation and progress of harmonious estimation theory in modern spectrum estimation are briefly reviewed, also the main works of this paper are introduced.In chapter 2, some statistics models of radar clutters are investigated under the high resolution radar, and some relation between the input and output correlation coefficient of zeros memory nonlinearity(ZMNL) are analyzed, then the fast numerical method(called error controlled method) which utilize calculating input and output correlation coefficient of ZMNL is proposed. After the mathematics principle which use simulating well and Guassian probability distribution are reviewed, the simulation method of correlation Gaussian clutter based on AR model is presented. Finally, some experiment results are given.In chapter 3, under the wideband radar system, some parameter estimation methods of scattering center are studied in the white noise and clutter background. Based on the multi scattering center theory, the frequency model describing electromagnetic scattering characteristics of target is developed. Then, some typical scattering center estimation methods are lucubrated. In clutter condition, the parameter estimation algorithm which is no sensitivity to clutter is studied. Finally, based on one-dimension structure characteristics of target, the fully polarized correlation match recognition algorithm using range profile and the correlation match recognition algorithm using the scattering center position are proposed, and the experiment results suggest that the two recognition algorithms have fine classification characteristics.In chapter 4, firstly, the two-dimension scattering model of target under high resolution radar system is developed, the two dimension resolution rates are reviewed and coordinate transformation of data are discussed. Then, in white noise condition, a two-dimension scattering center estimation method which is called matrix pencil method of two-dimension harmonious restructure (MP2D) is proposed. The two-dimension RELAX algorithm which is used estimating the two-dimensional scattering center in clutter circumstance is studied.In chapter 5, the main works of this dissertation are summarized and the future researched areas are pointed out.更多还原