空间微动目标ISAR成像若干技术研究

发布时间：2019-01-12 11:29

【摘要】：空间微动目标是一类重要的空间目标,随着人类空间活动愈加频繁,微动目标成像由于能够为目标识别和空间安全提供重要信息,已逐渐成为研究热点。然而这些目标由于大多不满足小角度近似,现有的成像手段难以达到较为理想的成像效果,所以还需作更深入的研究。本文针对微动目标成像的特殊性,在分析微多普勒和距离-慢时间回波的基础上对成像难点和成像方法进行了概述。优化了基于非微动参量估计的时频分析瞬时成像法。从提高微动参数误差容限的角度改进了后向投影变换成像法,并将其推广到了进动旋转对称目标上去。从在固有微动参数误差容限下提高成像效果的角度改进了基于重采样和压缩感知的自旋目标成像法。本文主要工作如下:1.研究了自旋和进动目标的微多普勒和距离-慢时间回波的时变特性,分析了其成像的难点在于难以直接在方位向采用傅立叶变换压缩成像。根据是否需要事先估计自旋角速度等微动参数,将微动目标成像方法分为了基于非微动参量估计的成像法和基于微动参量估计的成像法,并简要介绍了其处理过程。2.研究了基于非微动参量估计的自旋目标时频分析瞬时成像法,针对目标转速较大时随机选取瞬时像可能选取到成像质量较差的图像来这一不足,提出了一种分段优选策略。采用图像对比度最优准则,最终得到一组不同时间段下的较好目标图像,并进行了仿真验证。3.分析了在有角速度估计误差时的自旋目标成像,基于微动参量估计的后向投影变换法(BPT)主要表现为能量发散并出现虚假点,此外当存在网格误差也会出现虚假点。针对这些不足,在分析包络检测的基础上提出了一种结合广义Radon变换(GRT)的联合成像法(GRT-BPT)。该方法利用加权互筛在成像效果和误差容限间取得了一个平衡,推广到进动旋转对称目标成像中也取得了较好效果。4.从在固有微动参数误差容限下提高成像效果的角度研究了角速度存在估计误差时的快速自旋目标成像法,在目标回波传统周期重采样消除自旋效应的基础上提出了自旋对称目标半周期重采样法以及遮挡目标间隔半周期重采样法。并对重采样所需自旋角速度估计精度进行了分析,指出了其误差容限。由于重采样后得到的回波数据较少,文中改进了标准傅立叶基,并利用压缩感知手段成像。仿真实验表明其避免了重采样后直接进行RD成像的图像模糊问题。
[Abstract]:Space fretting target is an important kind of space target. With the increasing of human space activities, fretting target imaging has become a research hotspot because it can provide important information for target recognition and space security. However, because most of these targets do not satisfy the small angle approximation, the existing imaging methods are difficult to achieve a more ideal imaging effect, so more in-depth research is needed. Aiming at the particularity of fretting target imaging, this paper summarizes the imaging difficulties and imaging methods based on the analysis of micro-Doppler and range-slow time echo. The time-frequency analysis instantaneous imaging method based on non-fretting parameter estimation is optimized. In this paper, the backward projection transform imaging method is improved from the angle of increasing error tolerance of fretting parameters, and it is extended to precession rotating symmetric target. The spin target imaging method based on resampling and compression sensing is improved from the point of view of improving the imaging effect under the error tolerance of inherent fretting parameters. The main work of this paper is as follows: 1. The time-varying characteristics of micro-Doppler and range-slow time echo of spin and precession targets are studied. The difficulty of imaging is that it is difficult to use Fourier transform compression imaging directly in azimuth direction. According to whether it is necessary to estimate the spin angular velocity and other fretting parameters in advance, the fretting target imaging method is divided into two methods: the imaging method based on the non-fretting parameter estimation and the imaging method based on the fretting parameter estimation, and the processing process is introduced briefly. 2. In this paper, the time-frequency analysis instantaneous imaging method of spin target based on non-fretting parameter estimation is studied. In order to solve the problem of random selection of instantaneous image to the image with poor imaging quality when the speed of target is high, a piecewise optimal selection strategy is proposed. Finally, a group of better target images in different time periods are obtained by using the optimal contrast criterion, and the simulation results are verified. In this paper, the spin target imaging with angular velocity estimation error is analyzed. The backward projection transformation method based on fretting parameter estimation (BPT) is mainly characterized by energy divergence and false points, in addition, there are false points in the presence of mesh errors. Based on the analysis of envelope detection, a combined imaging method (GRT-BPT) combined with generalized Radon transform (GRT) is proposed. A balance between the imaging effect and the error tolerance is obtained by using the weighted mutual sieve, and the method is extended to the imaging of the precession rotating symmetric target. 4. In this paper, the fast spin target imaging method with angular velocity estimation error is studied from the angle of improving the imaging effect under the tolerance of inherent fretting parameter error. Based on the traditional periodic resampling of target echo to eliminate the spin effect, a semi-periodic resampling method and a half-period resampling method for blocking the target are proposed. The accuracy of spin angular velocity estimation for resampling is analyzed and the error tolerance is pointed out. Because the echo data after resampling is less, the standard Fu Li leaf base is improved, and the compression sensing method is used to image it. The simulation results show that it avoids the image blur problem of RD imaging directly after resampling.
【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】：TN957.52

【参考文献】