基于知识的机载雷达杂波抑制技术研究
发布时间:2018-03-20 15:45
本文选题:空时自适应处理 切入点:协方差矩阵估计 出处:《电子科技大学》2017年硕士论文 论文类型:学位论文
【摘要】:杂波抑制是机载雷达目标探测面临的主要问题。空时自适应处理(STAP)算法对杂波进行空域和时域的联合处理,可显著提高机载雷达的杂波抑制能力。实际的杂波环境往往呈现出非均匀、非平稳特性,传统的采用与待检测单元相邻的样本对协方差矩阵进行最大似然估计的STAP算法性能受限,如何准确有效地估计待检测单元的杂波协方差矩阵,成为进一步提升机载雷达目标探测性能的关键问题。本文研究非均匀杂波环境下的空时自适应处理问题,分析了非均匀环境下进行数据样本筛选的必要性,对现有样本筛选算法进行了仿真,并利用实测数据验证了处理算法的实际性能;在此基础上,提出了一种改进样本自适应样本筛选算法,在小系统自由度情况下,具有更优的筛选性能。本文的具体内容概括如下:1.介绍机载雷达杂波回波模型及STAP基本理论。从全空时自由度处理的STAP算法出发,讨论了降维/降秩处理的必要性,介绍了常用降维STAP算法的基本原理,并对STAP处理性能的评估参数进行说明。2.对机载雷达的非均匀杂波环境,进行信号建模与杂波仿真。对比分析了杂波的最小方差功率谱与傅里叶谱。介绍了非均匀环境下的杂波协方差矩阵估计方法及知识辅助STAP(KA-STAP)滤波器的基本原理。3.研究了两种经典的独立同分布(IID)样本筛选算法,即广义内积(GIP)法和基于傅里叶谱相似度(FSPS)的样本筛选算法,讨论了不同的非均匀杂波场景下,两种算法的适用性,并进行了建模仿真和性能对比。结果表明,GIP算法对存在少量离散强杂波单元的非均匀环境有较好的筛选效果,但在存在大量非均匀样本的环境下,算法性能较差;FSPS算法则在强非均匀性杂波环境下呈现出良好的稳健性。利用两种算法对机载雷达实测数据进行了处理,验证了算法的有效性。4.针对在小系统自由度下,已有的FSPS算法在污染样本剔除及相似样本选择环节都存在分辨率不足的问题,提出一种基于稀疏恢复技术的自适应样本筛选算法。该方法利用高精度稀疏恢复信息对参考单元样本进行筛选。相比FSPS算法,该方法在小系统自由度情况下可同时提升污染样本剔除及相似样本选择时的性能。通过计算机仿真验证了该方法的有效性。
[Abstract]:Clutter suppression is the main problem in airborne radar target detection. The clutter suppression ability of airborne radar can be improved significantly. The traditional STAP algorithm using samples adjacent to the unit to estimate the covariance matrix has limited performance. How to estimate the clutter covariance matrix accurately and effectively is proposed. In this paper, the problem of space-time adaptive processing in heterogeneous clutter environment is studied, and the necessity of data sample selection in non-uniform environment is analyzed. The existing sample selection algorithm is simulated, and the actual performance of the algorithm is verified by using the measured data. On the basis of this, an improved sample adaptive sample screening algorithm is proposed, which can be used in the case of small system degrees of freedom. The detail contents of this paper are summarized as follows: 1. The clutter echo model of airborne radar and the basic theory of STAP are introduced. The necessity of dimension reduction / rank reduction processing is discussed based on the STAP algorithm of total space-time degree of freedom processing. This paper introduces the basic principle of commonly used dimensionally reduced STAP algorithm, and explains the evaluation parameters of STAP processing performance. 2. For the non-uniform clutter environment of airborne radar, The minimum variance power spectrum and Fourier spectrum of clutter are compared and analyzed. The estimation method of clutter covariance matrix and the basic principle of KAP KA-STAP filter in non-uniform environment are introduced. In this paper, two classical independent and distributed IID-based sample selection algorithms are studied. That is, the generalized inner product (GIP) method and the sample selection algorithm based on Fourier Spectrum similarity (FSPS). The applicability of the two algorithms in different heterogeneous clutter scenarios is discussed. The simulation and simulation results show that the GIP algorithm has a good screening effect on the non-uniform environment with a small number of discrete strong clutter elements, but in the presence of a large number of non-uniform samples, the GIP algorithm has a good selection effect on the non-uniform environment with a small number of discrete and strong clutter elements. The performance of the algorithm is poor and the FSPS algorithm shows good robustness in the environment of strong inhomogeneity clutter. Two algorithms are used to deal with the measured data of airborne radar, and the validity of the algorithm is verified. 4. In view of the small system degree of freedom, The existing FSPS algorithm has the problem of insufficient resolution in the selection of contaminated samples and similar samples. An adaptive sample selection algorithm based on sparse recovery technique is proposed. The method uses high precision sparse restoration information to filter reference unit samples. Compared with FSPS algorithm, the proposed algorithm is more efficient than FSPS algorithm. The proposed method can improve the performance of the proposed method for the removal of contaminated samples and the selection of similar samples at the same time in the case of small system degrees of freedom. The effectiveness of the proposed method is verified by computer simulation.
【学位授予单位】:电子科技大学
【学位级别】:硕士
【学位授予年份】:2017
【分类号】:V243.2;TN957.51
【参考文献】
相关期刊论文 前6条
1 刘汉伟;张永顺;王强;吴亿锋;;基于稀疏重构的机载雷达训练样本挑选方法[J];系统工程与电子技术;2016年07期
2 方明;刘宏伟;戴奉周;王小谟;;基于环境动态感知的空时自适应处理[J];电子与信息学报;2015年08期
3 吴亿锋;王彤;吴建新;代保全;同亚龙;;基于道路信息的知识辅助空时自适应处理[J];电子与信息学报;2015年03期
4 杨志伟;廖桂生;贺顺;曾操;;直接数据域的自适应匹配滤波方法[J];电子学报;2010年07期
5 周宇;张林让;刘楠;刘昕;;空时自适应处理中基于知识的训练样本选择策略[J];系统工程与电子技术;2010年02期
6 吴建新;王彤;索志勇;保铮;;空时自适应处理的滑窗递推QR算法[J];电子与信息学报;2008年10期
相关博士学位论文 前1条
1 阳召成;基于稀疏性的空时自适应处理理论和方法[D];国防科学技术大学;2013年
,本文编号:1639745
本文链接:https://www.wllwen.com/kejilunwen/xinxigongchenglunwen/1639745.html