【作者】 张绪锦；

【导师】 朱兆达；

【作者基本信息】 南京航空航天大学 ， 通信与信息系统， 2009， 博士

【摘要】 1971年R.Keith.Raney首次在他发表的论文中指出地面慢速运动目标对单天线合成孔径雷达(SAR)的图像质量有影响,从那时起30余年来各国学者对地面动目标检测问题进行了各个方面的研究,但真正认识到将SAR与地面动目标指示(GMTI)结合起来的重要性,应该是上世纪九十年代初,美军展示的E8A侦察样机JSTARS在海湾战争中的优异表现。从那时起,国内的高校和科研院所在各种背景项目的支持下,开展了不同程度的预研,取得了一定成果,但SAR/GMTI研究的实践性非常强,其理论成果需要大量的实测数据加以证实。在过去,由于国内没有这种体制的雷达,许多研究都是基于仿真数据的结果,或通过其它渠道获得的少量国外数据(如美国的MCARM数据),从而限制了研究的进一步深入。近年来,随着各种新体制SAR预研工作的开展,GMTI大多成为要求它们必有的功能之一,人们意识到对GMTI的研究工作不能再停留在纸上谈兵的阶段。在国内,华东电子工程研究所(电子38所)首次实现了多通道SAR/GMTI多模战场侦察雷达试验样机的飞行试验,录取了多种工作模式大量的实测数据,本文的研究工作正是利用该试验样机在GMTI模式下获得的数据进行的,各章内容概述如下:第一章是绪论,通过对国内外SAR/GMTI雷达发展历程的综述,揭示了动目标检测的基本问题;给出了本文的主要研究内容,论文的创新点和结构安排。第二章论述了相位中心偏置天线(DPCA)、自适应DPCA(ADPCA)、沿航迹向干涉(ATI)、杂波抑制干涉(CSI)和空时自适应处理(STAP)这几种适用于多通道接收动目标检测处理的方法,比较了它们各自的优缺点并对适用性作出分析。通过对地面动目标最小可检测速度(MDV)的分析,得出了双通道(天线)接收GMTI处理的MDV要比单通道的MDV小4倍的结论。第三章研究机载三通道SAR/GMTI在“广域GMTI”和“同时SAR/GMTI”模式下的信号处理方法。对处理过程中的ADPCA最佳权值的计算、通道间的相位补偿和杂波对消、信号的相参积累、动目标多普勒频率的估计、CFAR检测、动目标定位、径向速度估计、性能分析等环节进行了详细论述,对部分环节提出了符合工程应用的近似简化,并对近似引起的误差进行了可行性分析。针对影响地杂波对消和动目标定位精度的问题,处理时运用数据预处理、MTD处理和解动目标方位模糊三项关键技术予以解决。提出在杂波对消后采用基于有序统计的VI-CFAR器进行动目标检测,以消除GMTI对消后空间杂波剩余的分布不均匀和多普勒特性各不相同带来的影响。两种动目标检测模式均分别用计算机仿真验证了处理方法的有效性。第四章主要研究SAR/GMTI半实物仿真测试平台的构建方法,推导了仿真用的回波信号模型,用逐点叠加法和二维卷积法分别对点目标和面目标进行仿真;通过对半实物仿真测试平台采集到真实数据的分析,给出了对系统性能评估的内容、步骤、方法和结果。第五章详细、深入的研究了宽带系统I、Q通道校正和通道间幅相均衡校正的处理方法。从对点频信号的镜频分析入手,揭示了宽带线性调频(LFM)信号的镜像频谱具有对称性的特点,为了能有效地观察到线性调频信号的镜像频谱,本文提出了“时域切分”的观点,并据此给出了三种针对宽带雷达镜像处理和评估的方法,即幅度不变校正法、镜像切除法和归一化校正法。针对通道间幅相不均衡给GMTI处理带来不利影响这一棘手问题,论文结合实际多通道接收系统,提出了一套从雷达前端有源相控阵天线内的T/R组件到多通道接收机末级的完整解决方案,通过天线自动校正和通道间接收信号的均衡校正,在实测数据的处理中取得了良好的效果,所提供的通道均衡处理方法对InSAR、多波段SAR、多极化SAR、宽观测带SAR的通道均衡均具有较大的实用价值。第六章主要是处理从试验用三通道机载SAR/GMTI多模战场侦察雷达采集到的实测数据。提出了一种基于改进能量峰值法的多普勒中心估计,其目的是避免受强反射点目标的影响,而导致多普勒中心估计不准。实时处理的多普勒调频率估计则根据不同场景采用子图相关和最大对比度相结合的混合式方法。在SAR成像方面,为适应实时条带成像处理,对RD算法进行了必要的改进;结合试验雷达具有方位向电扫功能的特点,将聚束成像通过解调函数Dechirp处理后转为条带成像,通过增加扫描波位驻留脉冲的方式提高合成孔径时间内的分辨率。在GMTI处理方面,给出了对“同时SAR/GMTI”模式处理后获得三个通道的子图像,通道间对消后的残差子图像,对动目标检测处理中点迹过滤、凝聚和参数估计的方法和准则进行了论述,对合作目标的试验结果进行了定量分析,同时提供了对乡间公路、高速公路和铁路这些典型场景检测到的非合作动目标与SAR图像叠加后的结果。在“广域GMTI”对合作目标的ADPCA处理中,从处理结果观察到通过空间相位差补偿使差波束在和波束的峰点(或接近峰点处)形成凹口,达到了较好抑制地杂波的目的,经检测后形成多帧连续的动目标轨迹。论文的结束语部分,对本文的主要研究工作和结果作了全面总结,给出了一些与本文相关但尚待进一步深入研究的内容。更多还原

【Abstract】 In 1971, R.Keith.Raney first presented influence of slow ground moving target on quality of image formed by the synthetic aperture radar (SAR) with single antenna in his paper. During the last thirty years, scholars from different countries carried out various studies on ground moving target detection in different field. However, the essentiality to integrate the SAR with ground moving target indication (GMTI) function was not well accepted until the beginning of 1991 when US Airforce demonstrated the excellent performance of prototype of E8A battlefield reconnaissance radar JSTARS in Gulf war. Since then, domestic universities and research institutions had advance study on GMTI for different projects and made some achievements. Since the research on SAR/GMTI depends largely on practice, the theoretical results shall be proven with data obtained in the field test. In the past, there was no such kind of radar system in China. Many researches were based on simulated data, or data from foreign countries, such as Multi-Channel Airborne Radar Measurements (MCARM) from the United States. Therefore, the further study on GMTI was limited by the actual situation. In recent years, with advanced research on various SAR systems, GMTI has been required to be an essential function of them. So researchers realized that GMTI shall be put into field use. East China Research Institute of Electronic Engineering (ECRIEE) developed a new multi-mode SAR/GMTI battlefield reconnaissance radar which was firstly tested in China, large amount of data were acquired with the multi-channel SAR/GMTI multi-mode battlefield reconnaissance radar. The research work of this paper is mainly based on the data, the main contents of study are summarized as follows.Chapter 1 gives the brief introduction. It discusses and summarizes the general situation of SAR/GMTI techniques developed in the world, putting forward the basic problem of ground moving target detection. The main contents, structure and creativity of this paper are also presented.Chapter 2 describes different GMTI processing methods for multi-channel receiving system, such as DPCA, ADPCA, ATI, CSI and STAP. By comparing their advantages and disadvantages, the applicability is analyzed for each method. After deducing the minimum detectable velocity (MDV) of double-channel receivers system in detail, it comes to the conclusion that MDV of double-channel receivers system is one-fourth of single channel receiver system.Chapter 3 presents the studies on the signal processing technologies for three-channel airborne radar in wide-area GMTI and simultaneous SAR/GMTI modes. The processing methods such as calculation of optimum weight of ADPCA, phase compensate and clutter suppression between channels, signal coherent accumulation, Doppler estimation of moving target, CFAR detection, moving target location, radial velocity estimation and performance analysis are discussed in detail. Some processing methods are substituted for simplifying approximate expressions. Data pre-processing, MTD and unwrapping azimuth ambiguity of moving target are used to obtain better clutter suppression and locate moving target accuracy. An ordered statistics-based VI-CFAR detector is also used to remove the influence of non-uniform distribution and different Doppler frequency on space residual clutter. Computer simulation result shows that the moving target detection method presented in this chapter is valid.Chapter 4 contains approaches to establish a hardware-in-loop simulation test system for multi-channel SAR/GMTI. After deducing simulation echo model, the author uses point-to-point overlay and two dimensional convolution to simulate point target and regional target in the hardware-in-loop platform. By analyzing the real data collecting from the hardware-in-loop simulation test system, the contents, procedures, methods and results are given to evaluate the performance for new SAR/GMTI system in laboratory.Chapter 5 gives an in-depth study on channels I/Q calibration, equalizing amplitude and phase imbalance between channels. By analyzing different image frequency on narrowband frequency signal and linear frequency modulated (LFM) signal, the paper presents a new approach called“Time-domain division”to solve the image frequency of LFM. Based on the view of point, the author gives three new methods for processing and evaluating wideband image frequency, namely Amplitude Immovability Calibration, Image Removal and Normalized Calibration. To deal with the disadvantage of amplitude and phase imbalance between channels operating in GMTI processing, the paper puts forward a complete solving scheme for a multi-channel SAR/GMTI system. By self-calibrating T/R modules in active phased-array antenna and channels equalizing, the author got a good result in the processing of live data. Channels equalization technology presented in this chapter is very useful for wideband signal system, which is in the area such as InSAR, multi-band SAR, multi-polarization SAR and wide-swath SAR.Chapter 6 focuses on processing live data collected by three-channel SAR/GMTI system. To avoid strong reflect point target impact on estimate of Doppler centre frequency, an improved method based on energy peak value is given. According to different scenarios sub-image correlation and contrast maximization is combined for Doppler frequency modulation rate estimate in real-time processing. An improved Range-Doppler algorithm for real-time strip mode imaging. By increasing dwell pulses for each scan beam position high resolution in spotlight mode imaging is achieved. Spotlight imaging to strip imaging are realized with de-chirp demodulation function. A lot of processing results are given in simultaneous SAR/GMTI mode, such as sub-images of three channel, residual sub-images after channels cancellation. Method and rule of moving target plot filter, plot coherence and precise parameter estimation are discussed. The performances of cooperative moving targets are analyzed. Uncooperative moving targets added to SAR image are illustrated by processing some typical scenarios, such as countryside road, expressway and railroad. After space phase difference compensated in wide-area GMTI mode, it can be observed that△-beam forming a notch filter corresponding to peak energy point ofΣ-beam (or close to ) from the ADPCA processing result of cooperative target. So energy superpose forΣ-beam and energy is counteracted for△-beam. ADPCA processing principle is used to reach the purpose of desirable clutter suppression and forming continuous target tracks after detection.In the conclusion of this paper, the main contents and results are comprehensively summarized. Some questions associated with this paper are given as well, which deserves to be studied deeply in the future.更多还原