采用长时间相参积累技术的高速机动目标检测快速算法  

A Fast Detection Algorithm for High-Speed Maneuvering Targets Based on Long-Time Integration

作　　者：黄响 张林让[1] 唐世阳[1] HUANG Xiang;ZHANG Linrang;TANG Shiyang(National Laboratory of Radar Signal Processing, Xidian University, Xi’an 710071, China)

机构地区：[1]西安电子科技大学雷达信号处理国家重点实验室

出　　处：《西安交通大学学报》2019年第8期121-128,共8页Journal of Xi'an Jiaotong University

基　　金：国家自然科学基金资助项目(61601343,61731023,61671361)

摘　　要：针对雷达接收的高速机动目标回波能量较弱的问题,提出一种采用长时间相参积累技术的高速机动目标检测快速算法。首先,利用慢时间反转变换对速度和加加速度引起的三阶距离徙动和多普勒徙动进行同时校正;其次,根据搜索的加速度构造二次相位补偿函数,以校正加速度引起的二阶距离徙动和多普勒徙动,并实现目标能量的相参积累;最后,利用恒虚警技术对相参积累结果进行处理,以确定目标是否存在。与无需运动参数搜索的时间反转变换-二阶楔石变换-吕氏分布(TRT-SKT-LVD)算法相比,该算法简化了操作流程,避免了因插值运算造成的性能损失,且不受雷达回波冗余信息和运动参数估计范围的限制。仿真结果表明,与基于多维运动参数搜索的雷顿-分数阶模糊函数(RFRAF)和广义楔石变换-广义去调频(GKT-GDP)算法相比,该算法由于仅需对加速度参数进行搜索,因此计算复杂度降低近2个数量级,运算时间仅为原来的1/100。A new fast detection algorithm for high-speed maneuvering target utilizing the long-time coherent integration technique is proposed to solve the problem that the energy of radar echoes is weak for high-speed maneuvering targets. First, the slow time reversing transform is utilized to correct both the third-order range migration (RM) and Doppler frequency migration (DFM) induced by velocity and jerk. Then, a quadratic phase compensation function based on the searching acceleration is constructed to correct the second-order RM and linear DFM caused by the acceleration and to realize the coherent integration of the target’s energy. Finally, the constant false alarm technique is used to determine whether there exists a target. Compared with the method of time reversing transform, second-order Keystone transform and Lv’s distribution (TRT-SKT-LVD), the proposed method simplifies the operational processes and avoids the performance loss induced by the interpolation. Furthermore, the proposed method is not restricted by the redundant information of radar echo and the estimation range of motion parameters. Simulation results and comparisons with the Radon-fractional ambiguity function (RFRAF) method and the generalized Keystone transform-generalized Dechirp process (GKT-GDP) method based on multi-dimensional motion parameter search show that the proposed method reduces the computational complexity by nearly two orders of magnitude because of the only search of acceleration, and the computation time is only one percent of the original.

关 键 词：目标检测 相参积累 距离徙动 多普勒徙动 

分 类 号：TN957[电子电信—信号与信息处理]