窄带雷达中段目标识别技术研究

发布时间：2018-05-29 23:02

本文选题：弹道导弹防御 + 窄带雷达　；参考：《国防科学技术大学》2014年博士论文

【摘要】：深入挖掘窄带雷达潜力,研究基于窄带信息的弹道中段目标识别和特征提取技术对于弹道中段攻防对抗具有重要意义。本文以弹道导弹防御为背景,针对弹道中段目标识别难题,重点研究了基于轨道特征的有源假目标识别技术、基于窄带RCS(Radar Cross Section,RCS)序列的进动目标周期特征提取技术和基于时频变换域的进动目标成像技术。第一章阐述了课题研究背景及意义,简要介绍了弹道导弹攻防对抗的现状,对基于窄带雷达信息的弹道中段目标识别和特征提取相关技术进行了归纳总结和分析,最后介绍了本文的主要研究工作。第二章为弹道目标跟踪基础理论研究。首先介绍了跟踪滤波的基本原理,对基于动力学模型和基于运动学模型的弹道目标跟踪滤波的数学模型及关键要素进行了分析和讨论;然后从状态方程的预测性能和滤波过程的线性化两个方面,分别提出了基于运动学与动力学混合模型的弹道中段目标跟踪算法和基于变换观测量的雷达垂直坐标系下的弹道中段目标跟踪算法。所提方法兼具动力学和运动学两种模型的优点,即既具有与动力学模型相近的性能,又具有运动学模型线性化的结构。第三章研究了基于弹道中段目标轨道特征的有源假目标识别技术。首先深入分析了弹道目标和有源假目标的动量矩、近拱点矢量和轨道根数等轨道特征量。在此基础上,提出了基于轨道根数时不变特性的有源假目标识别技术,给出了时不变判定准则,基于轨道可逆性原理实现了初始观测数据的精确估计。最后对三种典型布站情况进行了仿真,仿真结果验证了所提方法的有效性。第四章研究了基于窄带RCS序列的弹道中段进动目标周期特征提取技术。首先利用电磁计算软件计算得到典型中段目标的RCS数据,通过样条拟合生成其动态RCS序列;简要介绍了传统的RCS特征提取方法,指出了传统方法在基于RCS序列提取进动目标周期特征时的局限性。然后从周期的定义和性质出发,分别提出了基于分组数据相似性度量和基于分组数据非参数统计特征的周期估计方法,给出了周期判定准则。仿真结果表明,所提方法能有效克服传统方法的缺陷。第五章研究了基于时频域的弹道中段进动目标窄带成像技术。首先概述了传统基于理想散射中心的进动目标窄带回波建模方法、时频分析技术和基于时频变换域的窄带成像方法,通过仿真分析指出了当存在非理想散射现象时传统成像方法的不足。然后建立了基于非理想散射中心的弹道中段进动目标窄带回波模型,提出在时频变换之前基于经验模态分解对信号进行分解,然后对各本征模态分别进行时频变换域成像的方法。最后通过暗室测量数据和暗室进动实验分别验证了模型的正确性和所提成像方法的有效性。第六章总结了论文的研究工作和主要创新点,指出需要进一步研究的问题。
[Abstract]:In order to exploit the potential of narrowband radar, it is very important to study the target recognition and feature extraction technology based on narrowband information. In this paper, based on ballistic missile defense, aiming at the difficult problem of target recognition in the middle of trajectory, the active false target recognition technology based on orbit feature is studied. The precession target periodic feature extraction technique based on narrowband RCS(Radar Cross sequence and the precession target imaging technology based on time-frequency transform domain. The first chapter describes the research background and significance of the subject, briefly introduces the current situation of ballistic missile attack and defense countermeasures, summarizes and analyzes the related techniques of target recognition and feature extraction in the middle part of trajectory based on narrowband radar information. Finally, the main research work of this paper is introduced. The second chapter is the basic theory research of trajectory target tracking. Firstly, the basic principle of tracking filtering is introduced, and the mathematical model and key elements of trajectory target tracking filtering based on dynamics model and kinematics model are analyzed and discussed. Then, the prediction performance of the equation of state and the linearization of the filtering process are discussed. The algorithms of midcourse target tracking based on the mixed kinematics and dynamics model and the radar midcourse target tracking algorithm based on the transformed observations in the vertical coordinate system are proposed respectively. The proposed method has the advantages of both dynamics and kinematics, that is, it has not only the same performance as the dynamic model, but also the linearized structure of the kinematic model. In the third chapter, the active false target recognition technology based on the trajectory characteristics of the middle trajectory is studied. First, the orbital eigenvalues such as moment of momentum, near arch point vector and orbital root number of ballistic target and active false target are analyzed. On this basis, an active false target recognition technique based on the time-invariant property of orbital root number is proposed, and the criterion of time-invariant decision is given, and the accurate estimation of initial observation data is realized based on the principle of orbit reversibility. Finally, the simulation results of three typical stations are given to verify the effectiveness of the proposed method. In chapter 4, we study the method of extracting the periodic feature of the moving target in the middle of trajectory based on narrowband RCS sequence. At first, the RCS data of typical middle target are calculated by electromagnetic calculation software, and its dynamic RCS sequence is generated by spline fitting, and the traditional RCS feature extraction method is introduced briefly. The limitation of traditional method in extracting precession target periodic features based on RCS sequence is pointed out. Then, based on the definition and property of the period, the method of period estimation based on the similarity measure of packet data and the non-parametric statistical feature of packet data is proposed, and the cycle criterion is given. Simulation results show that the proposed method can effectively overcome the shortcomings of traditional methods. In chapter 5, the narrowband imaging technology of the moving target in the middle part of trajectory based on time-frequency domain is studied. Firstly, the traditional narrowband echo modeling method based on ideal scattering center, time-frequency analysis technique and narrow-band imaging method based on time-frequency transform domain are summarized. The shortcomings of traditional imaging methods are pointed out by simulation analysis when there are non-ideal scattering phenomena. Then the narrowband echo model of the precession target in the middle trajectory based on the non-ideal scattering center is established and the signal is decomposed based on the empirical mode decomposition before the time-frequency transformation. Then the time-frequency transform domain imaging method is used for each intrinsic mode. Finally, the correctness of the model and the validity of the proposed imaging method are verified by the anechoic measurement data and the anechoic precession experiment, respectively. Chapter 6 summarizes the research work and main innovation points, and points out the problems that need further research.
【学位授予单位】：国防科学技术大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN957.52

【参考文献】