基于背景感知的合成孔径雷达动目标指示技术研究

项目名称： 基于背景感知的合成孔径雷达动目标指示技术研究

项目编号： No.61471185

项目类型： 面上项目

立项/批准年度： 2015

项目学科： 无线电电子学、电信技术

项目作者： 吕高焕

作者单位： 鲁东大学

项目金额： 55万元

中文摘要： 本项目研究基于背景认知的单天线合成孔径雷达地面动目标指示理论，包括动目标检测算法、目标运动状态估计理论，并对所提算法进行性能评价。动目标检测包含合成孔径雷达图像分析、背景建模和动目标提取方法；运动状态估计理论包含距离向、方位向速度估计和方位向速度与距离向加速度的解耦方法。重点研究基于背景特征函数库的杂波抑制算法和用于检测动目标的对称多普勒两视法，基于鲁棒主成分分析的动目标参数提取方法，结合对称散焦滤波技术研究改进子孔径配准算法以提高动目标的检测性能和运动状态估计精度，探测综合利用自聚焦技术和运动学原理实现方位向速度和距离向加速度的解耦的有效途径。通过建立统计模型以评价算法性能，包括评估给定恒虚警条件下的检测概率以及不同目标背景比条件下的最小可探测速度；通过理论分析和实验评估不同目标背景比条件下的速度估计精度。

中文关键词： 合成孔径雷达；地面动目标指示；速度估计；自聚焦

英文摘要： The background cognition based ground moving target indication algorithms will be researched by using a single-antenna synthetic aperture radar (SAR) in this project. Moving target detection algorithms consist of the SAR image analysis, background modeling method, and moving target extraction algorithms. Moving status estimation methods include the range and azimuth velocity component estimation algorithms, and methods for decoupling azimuth velocity and range acceleration components. The research works put emphasis on the clutter suppression algorithm which is based on background feature function library, the symmetric Doppler based two-views method for moving target detection, the robust principal component analysis method for moving target parameters extraction, and the improved subaperture registration algorithm combined with symmetric defocusing filters for improving the detection performance and estimation accuracy, and algorithms which are used to decouple the azimuth velocity and range acceleration combining autofocusing algorithms and kinematics thoery. The performance of the proposed algorithms, including the detection probability, the minimum detectable velocity and estimation accuracy under different target-to-background-ratio conditions, will be evaluated by establishing statistical models, theoretical analysis, and experiments.

英文关键词： synthetic aperture radar;ground moving target indication;velocity estimation;autofocus