弹载双天线前视SAR成像算法研究

发布时间：2018-04-30 02:22

本文选题：合成孔径雷达 + 前视成像算法　；参考：《南京理工大学》2017年硕士论文

【摘要】：合成孔径雷达(Syntheyic Aperture Radar,简称SAR)是一种能够全天时全天候工作的成像雷达,可以在能见度极低的气象条件下得到类似光学照相的高分辨雷达图像,所以被广泛地使用。将SAR成像技术与导弹导航相结合,能够利用SAR图像信息与惯性导航系统定位信息相比较,校对导航误差,提高打击效率。但是弹载平台的SAR有前视、非匀速直线运动等特点,这使得传统的成像方法无法获得高分辨率的雷达图像,所以本文结合导弹的运动特性,分别从平飞弹道和俯冲弹道两种运动轨迹对弹载前视SAR成像方法进行了研究。在平飞弹道下,弹载前视SAR由于多普勒频率变化小,所以存在成像盲区,并且飞行方向两侧会出现具有相同多普勒历程的对称目标,导致成像出现左右模糊问题。本文首先根据平飞弹道下弹目间的几何关系,推导出弹载前视SAR成像的目标回波信号模型,分析了时域下瞬时目标斜距的泰勒级数展开式以及式中各项对前视成像的影响;然后提出一种适用于平飞弹道下基于双天线的弹载前视SAR的改进距离多普勒成像算法,并且利用双天线模式下不同天线信号间的波程差,来解决左右模糊问题;最后通过仿真实验验证了算法的可行性。在俯冲弹道下,由于弹目间的相对运动导致回波信号发生严重的距离徙动现象,还会使得多普勒调频斜率不断变化,引起方位向严重的聚焦深度问题,成像的分辨率大大降低。通过分析俯冲弹道下弹载平台的运动特点,构建其前视SAR成像的目标回波信号模型,并推导出二维回波信号的表达式;然后从频域算法的角度出发,提出一种适用于双天线单平台前视SAR成像系统的改进非线性频调变标成像算法,采用线性距离走动校正、距离徙动校正和二次距离压缩完成距离向与方位向的解耦合和距离向压缩;接着采用非线性频调变标操作校正多普勒频率的空变性完成方位向压缩;最后通过仿真实验验证了算法是否可行。本文针对平飞弹道与俯冲弹道的前视SAR,提出各自适用的成像算法,通过仿真实验证明了成像算法的可行性,具有一定的理论参考价值。
[Abstract]:Synthetic Aperture Radar (SAR) is a kind of imaging radar which can work all day and all weather. It can get high-resolution radar images similar to optical photography under extremely low visibility, so it is widely used. The combination of SAR imaging technology and missile navigation can make use of the SAR image information to compare with the positioning information of inertial navigation system to correct the navigation error and improve the strike efficiency. However, the SAR of the projectile platform has the characteristics of forward view and non-uniform linear motion, which makes the traditional imaging method unable to obtain high-resolution radar images. So this paper combines the motion characteristics of the missile. The forward looking SAR imaging method of projectile is studied from two kinds of moving tracks of flat missile track and subduction trajectory respectively. Under the plane missile track, there is a blind area of imaging because of the small change of Doppler frequency, and the symmetrical target with the same Doppler course will appear on both sides of the flight direction, which leads to the problem of left and right blur in the imaging. In this paper, the target echo signal model of projectile forward SAR imaging is derived according to the geometric relationship between projectiles under the plane missile track, and the Taylor series expansion of instantaneous target oblique distance in time domain and the influence of each item in the expression on forward imaging are analyzed. Then an improved range Doppler imaging algorithm based on dual antennas for projectile forward looking SAR is proposed and the problem of left and right ambiguity is solved by using the wave range difference between different antenna signals in dual antenna mode. Finally, the feasibility of the algorithm is verified by simulation experiments. Under subduction trajectory, the relative motion between projectiles and objects leads to a serious range migration phenomenon of echo signal, and also makes the Doppler frequency modulation slope change continuously, causing a serious azimuth focusing depth problem, and the resolution of imaging is greatly reduced. By analyzing the motion characteristics of projectile platform under subduction trajectory, the target echo signal model of forward looking SAR imaging is constructed, and the expression of two-dimensional echo signal is deduced, and then from the point of view of frequency domain algorithm, An improved nonlinear frequency-modulated scalar imaging algorithm for dual-antenna single-platform forward-looking SAR imaging system is proposed, which adopts linear range walk correction. Range migration correction and secondary range compression complete the decoupling and range compression of range and azimuth direction, and then the null variation of Doppler frequency is corrected by nonlinear frequency modulation scaling operation to complete azimuth compression. Finally, the feasibility of the algorithm is verified by simulation experiments. In this paper, aiming at the forward looking SARs of the flat-missile track and the subduction trajectory, the imaging algorithms are proposed. The simulation results show that the imaging algorithm is feasible and has certain theoretical reference value.
【学位授予单位】：南京理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN957.52

【参考文献】