节点文献

单脉冲雷达目标三维成像与识别研究

Study of Monopulse Radar Target Three Dimensional Imaging and Recognition

【作者】 李强

【导师】 张守宏;

【作者基本信息】 西安电子科技大学 , 信号与信息处理, 2007, 博士

【摘要】 雷达成像在精密制导、目标识别、民航管制等领域有着广泛的应用。逆合成孔径雷达(ISAR)可以得到目标的距离-多普勒像,但是在目标姿态变化时,ISAR像不能够反映目标的真实形状。对回波的各个分辨单元进行单脉冲测角,将方位、俯仰和距离信息结合起来就得到目标的三维像,此像与目标的物理尺寸一致,并且受目标的机动影响较小,十分有利于改善识别性能,但是单脉冲三维成像存在三维运动补偿、角闪烁等问题。本文主要论述单脉冲三维成像与目标识别方法,对抑制角闪烁的单脉冲测角方法、海杂波背景下的目标三维成像和目标三维像自动识别方法作了深入地研究。具体工作概括如下:1.当单脉冲雷达与目标距离很近时,角闪烁将成为主要测角误差源,其大小可能使测角结果指向目标尺寸以外。因此必须对角闪烁进行有效抑制,以提高测角精度。我们首先分析了由于目标切向运动引起的差波束方向图调制问题,推导出在对差通道回波信号进行高分辨处理后,谱展宽程度与目标运动参数之间存在某种关系,从而提出一种新的拉伸体制下的角估计方法。该方法通过搜索高分辨距离像波形熵的全局最大值得到目标角度估计,可以很好地抑制测角闪烁的发生。结合该测角方法,我们还给出了基于差波束ISAR像的角运动参数估计与补偿方法。2.无论是对目标的检测、成像或者识别,杂波均会造成严重地影响。以雷达对海面目标进行观测为例,当擦海角(入射余角)接近90°时,海面的散射系数σ0会增大到+10dB左右,此时回波距离像中目标将被杂波完全淹没。我们对这种情况下的海杂波抑制方法进行了探索性地研究。首先利用ISAR技术,对海面和目标同时成像。然后利用新提出的基于Radon变换的目标杂波分离算法将ISAR像中的海杂波抑制掉,再对目标散射点进行单脉冲测角。这样就使得雷达能够在高擦海角时获得目标的三维像。3.本文针对如何利用雷达三维成像技术实现自动目标识别的问题进行了探索。我们首先通过CAD(计算机辅助设计)技术建立目标模板库。在得到目标的三维像之后,估计出目标的姿态,修正已经建立好的目标模板库,通过与形态滤波相结合的最大融合度分类器完成目标的识别。最后仿真了三维目标识别方法,并对结果进行了分析。本论文各章节安排如下:第一章介绍了雷达成像与目标识别的研究背景及意义,回顾了前人的研究成果,并简要介绍了本文的研究重点。第二章分别介绍了距离高分辨技术、ISAR成像技术以及基于ISAR的单脉冲三维成像技术的基本原理与方法。第三章重点研究了能够有效抑制角闪烁的单脉冲测角方法。提出了拉伸体制下的最大熵测角方法,同时给出了基于差波束ISAR像的角运动参数估计与补偿方法。第四章针对高擦海角观测条件下的海杂波抑制方法进行了研究,提出了基于Radon变换的杂波与目标分离方法,可以在ISAR像中将海杂波抑制掉,最终获得目标的三维像。第五章主要讨论的是如何利用基于拉伸处理的数字波束形成技术在宽带相控阵上实现三维成像功能。仿真对比了步进跟踪与连续跟踪两种情况下宽带相控阵对空中目标的单脉冲三维成像。第六章重点讨论目标三维像的识别方法。对模板库的建立、目标三维像的姿态估计以及基于形态滤波的最大融合度分类方法均作了详细论述。第七章对全文作了总结,展望了雷达目标三维成像与识别技术的发展方向及存在的一些问题。

【Abstract】 Radar imaging is widely applied in the fields of precision guidance, target recognition and air traffic control. But the range-Doppler image of Inverse Synthetic Aperture Radar (ISAR) cannot capture the true shape of maneuvering target. Monopulse three-dimensional imaging, reconstructed from three-dimensional spatial coordinates (range, pitch, azimuth) of scatter points of the target measured by monopulse radar, is consistent with the real size of target and less sensitive to maneuvering motion and very useful in improving the quality of recognition. However, problems such as three-dimensional motion compensation and angle glint still exist. This thesis addresses monopulse three-dimensional imaging technique and target recognition methods. The research highlights method of angle glint suppression in monopulse angular measurement, target three-dimensional imaging in the background of sea clutter and automatic target recognition based on radar three-dimensional imaging.A brief description of the thesis is given as follows:1 .In monopulse angular measurement of short-range targets, angle glint will be the primary error source. The magnitude of errors may exceed the size of the target. Therefore, effective methods of angle glint suppression must be developed to improve the precision of angular measurement. Pattern modulation of difference beam echoes due to the tangential motion of target is investigated firstly. Relations between the spectrum width of high-range-resolution profile of difference beam echoes and angle motion parameter of target are analyzed then. A novel method of angle motion parameters estimation is presented. It can suppress the angle glint effectively since the angle motion parameter of target is obtained through searching the global maximum entropy of the high-range-resolution profile. An extended method of angle motion estimation and compensation based on the ISAR imaging for difference beam echoes is presented subsequently.2.Clutter is a serious problem in target detection, imaging and recognition. As airborne radar observes targets on the sea surface at high grazing angle approaching 90°, scatter coefficientσ0 becomes so great (around +10dB) that targets will be completely submerged by sea clutter in the range profiles of echoes. This paper reports an exploratory development on sea clutter suppression in such condition. Firstly, ISAR imaging is applied to get the image of target in the sea clutter background. Secondly, the sea clutter in ISAR image plane is suppressed by the proposed method based on Radon transform. Finally, the monopulse three-dimension imaging method is applied. In this way, the effects of sea clutter at high grazing angle are eliminated and the monopulse three-dimension image of target on the sea surface can be obtained.3.This thesis presents an exploratory development on the automatic radar target recognition method utilizing monopulse three-dimension imaging technique. Computer Aided Design (CAD) technique is used for constructing the template of target firstly. Templates are then modified depending on the target’s attitude estimated from monopulse three-dimension image. Target recognition is accomplished by a maximal degree of fusion classifier with morphological filter. The process of three-dimension target recognition is simulated in the end.The dissertation is organized as follows:Chapter 1 provides the background, scientific significance of radar target imaging and recognition, a literature review of previous work and an overview of this research.Chapter 2 describes methods of high-range-resolution processing, ISAR technique and principle of ISAR based monopulse three-dimensional imaging.Chapter 3 in particular presents an effective method for angle glint suppression using monopulse angular measurement. It also provides a novel method of high-range-resolution profile based on maximum entropy, together with an extended method of angle motion estimation and compensation based on the ISAR imaging for difference beam echoes.Chapter 4 describes research methods about sea clutter suppression in target three-dimension imaging at high grazing angle. A novel method based on Radon transform to eliminate the effects of sea clutter is presented.Chapter 5 provides methods of three-dimensional imaging in wide-band phased array radar. Simulations of dechirping and DBF (digital beam forming) technique based monopulse three-dimensional imaging in two tracking modes ( step and continuous mode) in wide-band phased array radar are conducted.Chapter 6 describes research methods about target recognition utilizing monopulse three-dimension imaging technique. Methods of template modification, attitude estimation and target classification are detailed.Chapter 7 provides a summary of the dissertation, unsolved problem and development of monopulse target three-dimensional imaging and recognition for future research.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络