旋转式合成孔径雷达三维成像方法研究

发布时间：2018-03-08 02:36

本文选题：旋转式合成孔径雷达　切入点：三维成像　出处：《西安电子科技大学》2014年博士论文　论文类型：学位论文

【摘要】：旋转式合成孔径雷达(ROSAR)是一种新型的雷达成像模式，它既保留了传统合成孔径雷达(SAR)全天候、全天时、电磁穿透等优点，又具有重访周期短、全视域成像等卓越性能，广泛应用于自然灾难应急救援、地下资源勘探、公共场所无损安检、战场监视以及低空火力支援等领域。直升机等旋翼飞行器是ROSAR模式最为典型的应用平台，ROSAR天线固定在刚性支架上，并指向外围随机翼一起做匀速圆周运动，天线旋转一周即能完成周围场景一次探测。该技术巧妙地利用了机翼旋转特性，仅需要一个天线的旋转运动，就能形成方位向圆弧形合成孔径，从而实现周围场景二维成像，极大地提高了飞行器的飞行安全。但是面对低空飞行任务，周围场景通常非常复杂，而获取的二维图像将可能错误地反映潜在威胁物的空间信息，为此开展ROSAR三维成像研究显得尤为必要。本文以满足低空空域安全飞行为目的，针对ROSAR三维成像中的关键问题和技术难点，围绕973项目“复杂低空飞行的自主避险理论与方法研究”、国家自然科学基金“机载毫米波雷达旋转合成孔径成像处理方法研究”等项目的研究任务，从ROSAR三维成像模型和三维成像方法等方面展开研究。具体的工作和贡献如下： 1.介绍了传统的二维ROSAR模型，为后续的三维成像理论奠定基础。针对现有的二维成像方法运算量大、大方位角散焦问题，提出了一种基于频谱重构的二维ROSAR成像方法。通过分析ROSAR回波信号形式，对其直接进行傅立叶变换，获取二维精确频谱表达式。由于精确频谱的形式过于复杂，导致后续处理无法进行，折中考虑频谱的精确性和可操作性，采用四阶频谱重构的方法，获取形式简洁、保留较多信息的高阶频谱，并基于该频谱提出ROSAR距离徙动算法(RMA)和调频变标算法(CSA)。仿真实验表明，在方位角不大于90度情况下，频谱重构方法都能实现高精度快速成像。 2.建立了ROSAR干涉成像模型，并重点分析了非理想情况载机平台的轴偏移问题。通过ROSAR技术和干涉方法相结合，要求载机平台在不同高度上进行悬停，每悬停一次获取一组场景数据，对录取的数据进行成像和干涉处理，可实现周围场景全视域三维成像。针对平台运动非理想情况，讨论了轴偏移对方位带宽、图像位置和干涉相位的影响。轴偏移导致干涉相位中增加一项附加相位，通过推导和分析附加相位形式，构建了一个相位补偿函数，由于该函数需要场景高度这一未知信息，预先假设所有场景高度为零，然后通过构建相位加权因子来消除高度置零的影响。最后对轴偏移的影响进行仿真实验，，并获取了仿真场景三维高程图。 3.建立了旋转上升合成孔径雷达(SSAR)三维成像模型，并提出一种适用于该模型的三维成像算法。在现有的二维ROSAR技术的基础上，SSAR利用载机平台上升在高度向上形成第二个合成孔径，天线受水平旋转和匀速上升的共同作用，在空中形成一个圆柱形合成阵面，同时结合距离向发射的宽带信号，使得该模型具有三维成像能力。关于三维成像算法，为了降低成像处理难度，首先分析方位向采样和高度向上升对天线位置的影响，并构建位置偏移补偿函数，将“旋转上升”模式简化为沿高度向“一步一悬停”模式。然后分析了简化模式的相位历程，计算高度向中心频率偏移量，推导波数域三维匹配函数，构建波数域插值函数，进而实现简化模式的三维成像。最后对该模型进行了性能分析，并通过仿真实验验证成像算法的有效性。 4.建立了前行ROSAR(FMROSAR)三维成像模型，并针对该模型提出了一种三维成像算法。在前行状态下，天线指向前方区域发射宽带信号，同时受到平台前行和机翼旋转的共同影响，在空中形成一个水平合成阵面，结合距离向高分辨能力以及二维阵面分辨能力，FMROSAR可实现三维前视成像。FMROSAR成像算法包括模式简化和简化模式三维成像两大步骤。首先通过方位向偏移补偿和顺轨向偏移补偿，将前行模式简化为沿顺轨向“一步一停”模式。针对“一步一停”模式，又细分为距离-方位成像和距离-顺轨成像两个子步骤。在后者的成像过程中，分析了平台运动对方位角变化的影响，并给出了顺轨向有效聚焦的约束条件。同时考虑顺轨向大斜视成像情况，提出了一种改进的距离多普勒算法(RDA)。最后分析了上述模型的性能，并验证了成像算法的有效性。
[Abstract]:Rotary synthetic aperture radar (ROSAR) is a new radar imaging mode, it retains the traditional synthetic aperture radar (SAR) all-weather, all day long, and has advantages of electromagnetic penetration, revisit period, full horizon imaging, superior performance, are widely used in natural disaster emergency rescue, underground resources exploration. Public security surveillance and nondestructive, low fire support and other fields. Helicopter rotor aircraft is the most typical application platform for ROSAR mode, the ROSAR antenna is fixed in a rigid frame, and point to the periphery of the random wing do uniform circular motion, antenna rotation can be completed around the scene of a detection. This technology skillfully the wing rotation characteristics, rotational motion requires only one antenna, can form azimuth circular synthetic aperture, so as to realize the scene around two-dimensional imaging, greatly improve the vehicle Flight safety. But in the face of low altitude missions, the surrounding scenes are usually very complex. The two-dimensional images acquired will probably reflect the spatial information of potential threats wrongly. Therefore, it is necessary to carry out ROSAR 3D imaging research.
In this paper, in order to meet the low altitude airspace flight safety for the purpose, aiming at the key problem of ROSAR three-dimensional imaging and technical difficulties, around the 973 project "complex low flying self hedging theory and research method, research project of National Natural Science Foundation of airborne imaging rotating synthetic aperture millimeter wave radar research method", and carries out the research on ROSAR 3D imaging model and 3D imaging methods. The specific work and contributions are as follows:
The 1. introduces the traditional two-dimensional ROSAR model, lay the foundation for the subsequent three-dimensional imaging theory. According to the calculation of two dimensional imaging method of existing large amount of large azimuth defocus problem, proposed a two-dimensional ROSAR imaging method based on spectral reconstruction. By analyzing the form of the ROSAR echo signal and the Fu Liye transform directly to obtain two-dimensional the precise spectrum expression. Due to the precise spectrum form is too complex, resulting in subsequent processing can not be carried out, the trade-off between accuracy of spectrum and maneuverability, by using the method of four order spectrum reconstruction, obtain simple form, high order spectrum retain more information, and based on the proposed ROSAR spectrum range migration algorithm (RMA) and FM scaling algorithm (CSA). Simulation results show that the azimuth angle of not more than 90 degrees, the spectrum reconstruction method can achieve fast imaging with high precision.
2. an imaging model of ROSAR interference, and analyzes the problem of non ideal axis aircraft platform. Through the combination of ROSAR technique and the interference method, requirements of aircraft platform were hovering at different heights, each one gets a set of hovering scene data, on admission according to the number of imaging and interferometric processing. Can realize the full view around the scene of 3D imaging. Based on the ideal of non platform movement, discusses on the azimuth bandwidth axis offset, affect the image position and the interference phase. The shaft offset resulted in increased an additional phase interferometric phase, through the derivation and analysis of the additional phase form, construct a phase compensation function, because the function the height of the unknown scene information, presupposes all the scenes height is zero, then through the construction phase weighting factor to eliminate the influence of height zero. Finally simulation effect on axial offset The experiment was carried out and the 3D elevation map of the simulation scene was obtained.
3. the establishment of a spin up of synthetic aperture radar (SSAR) imaging model, 3D imaging algorithm and propose a suitable for this model. Based on two-dimensional ROSAR technology on the SSAR based aircraft platform increased form second in height to the synthetic aperture, the interaction of antenna by horizontal rotation and uniform rising and form a cylindrical synthetic array in the air, combined with the distance to transmit broadband signal, makes the model has three-dimensional imaging capabilities. A 3D imaging algorithm, in order to reduce the difficulty of image processing, the first analysis of azimuth sampling and height to the impact of rising position of the antenna, and construct the offset compensating function, simplified spin up "mode to" step along the height of a hover mode. And then analyzes the simplified phase process model, calculating the height to the center frequency offset, derived 3D wavenumber domain The matching function is used to construct the interpolation function in wavenumber domain, and then the simplified mode of 3D imaging is realized. Finally, the performance of the model is analyzed, and the effectiveness of the algorithm is verified by simulation experiments.
4. established ROSAR before (FMROSAR) 3D imaging model, and presents a 3D imaging algorithm based on this model. In the moving state, the antenna pointing to the front region to transmit broadband signals, at the same time by the platform forward and wing rotating joint effects of the formation of a water level synthesis array in the air, with high range resolution two dimensional array capability and resolution, FMROSAR can realize three-dimensional forward-looking imaging.FMROSAR imaging algorithm including model simplification and model of 3D imaging of two steps. Firstly, the azimuth offset compensation and orbit to offset compensation, will walk along the track model is simplified as "a step in a stop mode. According to the" one step one stop "mode, and subdivided into range azimuth imaging and imaging distance along track two sub steps. In the imaging process, analyzes the influence of platform motion azimuth changes, and gives the travel direction Considering the constraint condition of effective focusing, an improved range Doppler algorithm (RDA) is proposed considering the situation of imaging along the track to the large squint. Finally, the performance of the above models is analyzed, and the effectiveness of the imaging algorithm is verified.

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

【参考文献】