合成孔径雷达分辨率增强方法研究

发布时间：2018-01-28 03:30

本文关键词： 合成孔径雷达 RD算法频谱外推相干处理多频段合成　出处：《西安电子科技大学》2014年硕士论文　论文类型：学位论文

【摘要】：合成孔径雷达(Synthetic Aperture Radar,SAR)具有全天时、全天候、远距离成像的特点。提高分辨率是SAR领域的重要目标,随着SAR技术领域的推广,要求SAR系统可在不同体制和工作模式下具有高分辨成像能力,SAR的工作模式和相应的成像算法是实现高分辨成像的主要方面,但是随着分辨率要求的提高,常规的SAR成像算法很难达到高分辨要求,论文就高分辨SAR成像算法进行研究,通过仿真实验验证算法的可行性,主要研究工作包括:1、详细介绍了距离徙动概念,在此基础上推导了经典距离多普勒(RD)算法和校正线性距离走动的距离多普勒算法,讨论了两种算法的优势和局限性,并使用点目标成像实验验证算法的可行性,最后简单介绍了几种常用的成像算法,如CS,NCS算法,给出具体算法流程。2、提出一种基于频谱外推的SAR成像分辨率增强方法。将目标的极化信息与频谱外推技术相结合:首先采用Pauli分解方法将全极化数据投影到反映三种不同散射机理的Pauli基上;然后利用线性预测模型和目标不同极化通道数据间的冗余性与互补性,推导出全极化模式下的频谱外推算法;利用三路全极化数据对任意一路极化数据进行频谱外推,提高了频谱外推的精度。相比传统的频谱外推法,由于采用了联合外推的方法,能够获得更高的分辨率。3、提出一种实现超宽带信号的方法。带宽越大分辨率越高,但是超宽带信号很难实现,在产生和处理上会带来很大的硬件压力,论文提出一种通过对多个频段信号进行处理,得到一个等效超宽带信号的方法,该方法通过对信号建立全极点模型来实现,首先利用奇异值分解来估计子带的模型阶数,使用模型阶数将信号分为信号加噪声子空间和噪声子空间,利用改进的root-MUSIC算法来估计子带的信号极点;然后利用线性最小二乘法对子带的极点模型系数进行估计,对子带模型进行调整;最后利用非线性最小二乘法对合成整体频带全极点模型参数进行估计,为了验证子带互相干处理和全极点模型估计算法在提高分辨率方面的作用,进行SAR成像实验,得出理想的成像效果。
[Abstract]:Synthetic aperture radar (Synthetic Aperture, Radar, SAR) has all day long, all-weather, long-range imaging characteristics. To improve the resolution is an important goal in the field of SAR, with the promotion of SAR technology, SAR system has high resolution in different system and working mode, the working mode of SAR and the corresponding imaging algorithm is the main aspect of high resolution imaging, but as the resolution requirements for the improvement of conventional SAR imaging algorithm is difficult to meet the requirements of high resolution, the high resolution SAR imaging algorithm is studied through simulation experiment, the feasibility of the algorithm is verified, the main research work includes: 1, introduces the concept of range migration, on the basis of is the classical range Doppler algorithm (RD) and the linear range migration correction range Doppler algorithm, advantages and disadvantages of the two algorithms are discussed, and the use of point target imaging The feasibility of the algorithm is verified, and finally briefly introduces several commonly used imaging algorithms, such as CS, NCS algorithm, gives the algorithm of.2, a method is proposed to enhance the resolution of SAR image based on spectral extrapolation. The target polarization information and spectrum extrapolation technique combining decomposition method: first full polarimetric data to reflect the Pauli projection based on three different scattering mechanisms using Pauli; redundant and complementary and then using linear prediction model and target data between different polarization channels, spectrum extrapolation algorithm derived from the full polarization mode; spectral extrapolation of any circuit using polarization data of three channel full polarization data, improves the spectrum extrapolation accuracy. Compared with the traditional spectrum extrapolation method, because the method uses simultaneous extrapolation, can obtain higher resolution of.3, this paper puts forward a realization method of UWB signal bandwidth. The greater the resolution The higher the rate, but the UWB signal is very difficult to achieve, in the production and processing will bring great pressure to the hardware, this paper puts forward a based on multi band signal processing, get an equivalent method of UWB signals, the method to achieve the signal through the establishment of full pole model, firstly using singular value decomposition to estimate the model order with the use of the order of the model will be divided into signal and noise subspace signal plus noise subspace, to estimate the signal with the pole by using the improved root-MUSIC algorithm; then using pole model coefficients of linear least squares estimate subband, subband model adjustment; finally the synthesis parameters the whole band all pole model were estimated by nonlinear least squares method, in order to verify each sub-band processing and all pole model to estimate the effect of algorithm in improving the resolution of SAR The imaging experiment is used to obtain the ideal imaging effect.

【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN957.52

【相似文献】