【作者】 李军；

【导师】 邢孟道；

【作者基本信息】 西安电子科技大学 ， 信号与信息处理， 2011， 博士

【摘要】 逆合成孔径雷达(ISAR, Inverse Synthetic Aperture Radar)是一种全天时、全天候、远距离的微波探测工具,在国土防御、空间探测等领域发挥着越来越重要的作用。随着现代工业的发展及实际运用的需要,对ISAR成像技术的研究也在不断深入。本论文以二维成像至三维成像为主要发展路线,对高分辨二维和三维ISAR成像方法进行了研究。第一章介绍了本文的研究意义及背景,简要描述了ISAR成像的发展历史及研究现状。第二章介绍了高分辨ISAR成像的基本原理,对国内外现有的ISAR成像算法及研究成果进行了概述。第三章讨论了超宽带ISAR成像方法。从理论上分析了超宽带ISAR成像时的高阶近似问题。在一阶距离徙动校正的基础上,提出了利用DCFT变换估计二次相位,根据目标特显点拟合空变的调频率,然后对不同的距离单元分别进行相位补偿,方位FFT后,从而得到聚焦良好的二维高分辨ISAR像。本章最后对方位观测角度较大、目标散射特性发生变化的回波模型进行了初步讨论,提出了基于散射模型的目标分类方法。理论仿真和实测数据的处理验证了本章算法的有效性和可行性。第四章讨论了稀疏ISAR成像问题。针对ISAR系统中方位采样不均匀这一实际存在问题,构造了方位稀疏孔径下的ISAR成像模型,建立稀疏表征域及基空间表达式,通过分析稀疏情况下对方位聚焦的影响,提出了方位聚焦求解函数和约束条件,借助已有的优化算法对目标函数进行求解。考虑ISAR系统中回波信号非相参的影响,提出了基于柯西牛顿方法的多参数求解。针对实测ISAR数据,提出了稀疏孔径ISAR成像处理的一般流程,通过仿真和实测数据的处理对本章算法进行了验证。第五章研究了基于时频分析的高速自旋目标ISAR成像问题。在雷达发射窄带信号条件下,针对空间碎片、导弹等高速自旋目标,介绍了自旋情况下的雷达回波模型,提出了在时频域进行相干的单距离多普勒干涉算法,采用相干曲线积分对目标进行聚焦,相对传统的单距离多普勒干涉算法在图像分辨率方面有较大提高。然后分析了噪声及多普勒模糊对算法性能的影响,最后讨论了自旋速度误差对算法的影响及相应的精确估计方法。理论仿真和实测数据的处理验证了本章算法的可行性。第六章讨论了基于步进频率的自旋目标三维ISAR成像。首先在窄带子脉冲内部利用单距离匹配滤波算法对高速自旋目标进行二维ISAR成像,然后结合CLEAN算法确定目标散射点的个数及其极坐标下的散射点位置,提取目标散射点,最后在多个子脉冲二维像之间利用频带合成技术提高距离分辨率,从而实现三维ISAR成像。通过理论仿真验证了本章算法的可行性。第七章研究了基于雷达和差波束的三维ISAR成像方法。在和差波束测角雷达的基础上,增大发射信号带宽。首先利用多通道联合运动补偿方法对目标平动及初始相位误差进行校正,其次采用ISAR成像技术得到多个通道的距离多普勒像,去除背景干扰并提取有效目标散射点,再次利用比幅或者比相技术对单个通道ISAR像进行定标。最后通过多通道融合得到目标的三维像。理论仿真和实测数据的处理验证了算法的有效性。更多还原

【Abstract】 Inverse Synthetic Aperture Radar (ISAR), is an effective remote sensing tool in all-time and all-weather conditions, and has ability to provide an imagery of a target at very long range. ISAR has been playing more and more important roles in defense applications, such as surveillance of satellites and deep space targets. As the rapid development of industry and semiconductor, the advanced radar imaging technology has being studied. This dissertation focuses on study of some novel approaches to high resolution two-dimensional (2-D) and three-dimensional (3-D) ISAR imaging. The summary of this dissertation is given as follows:Chapter 1 is the introduction, which reviews the history of ISAR development and introduces the dissertation’s research background and related works.Chapter 2 presents the basic theory of high resolution ISAR. In this chapter, we review recent ISAR imaging algorithms, and some remaining problems are discussed breifly as well.Chapter 3 presents a novel approach to ultra-wideband ISAR imaging. First, a high-order approximation for instant range between target and radar is analyzed theoretically at the beginning of the chapter. Then DCFT (discrete chirp-fourier transform) is used to estimate the second-order phase error, which generates the migration through resolution cell. As the chirp rate is dependent on the range, the linear part can be obtained by searching the peak of the DCFT distribution. Subsequently, the second-order phase error can be compensated range by range. Finally, well-focused ISAR image can be generated by performming FFT in azimuth. In the last part of chapter, a backscatter model that has long coherent processing interval is presented. The new scattering characteristic is modeled as a function with respect to both the frequency and aspect angle, known as a parametric model. A method for target classification based on the scattering model is proposed. The validity is confirmed by real data and numerical simulations.Chapter 4 discusses the sparse aperture ISAR imaging technology. In the application of inverse synthetic aperture radar, the sparse data problem arises when continuous measurements are not possible or the measurements during certain periods are not valid. In this chapter, the interrupted or missing data model is described, and measurement matrix and dictionary are established. The thesis discusses a precise formulation of image formation as a sparse signal representation problem under a specific incomplete basis. Moreover a specific solution to the optimization function with regularization constraint is also introduced. Considering the phase error in the radar returns, a Quasi-Newton optimization is applied to minimize the cost function joint with retriving the phase error. Results with simulation data and real sparse ISAR data validate the feasibility and superiority of the approach.Chapter 5 studies the imaging processing for high speed spinning targets. It often occurs that the targets or some parts of a target are rotating along with the bulk translational motion, such as space debris, flying missiles, airscrews of airplanes, etc. The high-speed spinning turntable model with some assumptions is presented. Based on the time-frequency characteristic of the echoed signal, a novel coherent spectrogram redistribution method, coherent single range Doppler interferometry (CSRDI), is proposed, which is capable of generating high-resolution imagery by applying a phase matched filtering processing. In this chapter the cases of low SNR and Doppler Ambiguity are considered in detail. As the precision of spinning rate estimation is very important for both the existing approaches and CSRDI, the spinning rate error is taken into account and the estimation method based on image entropy is proposed in the last of chapter. The validity is confirmed by real data and numerical simulations.Chapter 6 studies 3-D ISAR imaging of spinning targets based on narrow-band stepped-frequency. A 3-D inverse synthetic aperture radar imaging method based on stepped-frequency waveforms for high speed spinning target is proposed in this chapter, by using narrow-bandwidth signal. Assuming the motion compensation is completed in advance, single-range matching filtering is performed in certain sub-pulse firstly to obtain ISAR image. Then parameters of scatters are estimated using CLEAN technique. ISAR images are obtained in different sub-pulse and different range bin. At last range synthetic profile processing is performed to get high resolution in range and the 3-D image is obtained. This method can void the Doppler influence of spinning targets on range synthetic and improve the resolution effectively. Simulations validate the feasibility and superiority of the approach.Chapter 7 discusses the 3-D ISAR imaging technology based on sum and difference beam. A wideband signal is transmitted in conditional angle-measurement monopulse radar. Firstly, the co-compensation for translational motion and phase error are performed in three receiver channels. Second, traditional processing is used to obtain the ISAR image and isolated scatters are chosen in image domain. At last the amplitude-comparison or phase interferometry with multi-channels is performed to scale the cross range bins. The results of real-data processing indicate that the algorithm proposed here performs better than traditional method, which validate the feasibility and superiority of the approach.更多还原