基于自适应预白化的水下动目标LFM回波联合检测  

Joint detection algorithm for underwater moving target's LFM echo based on adaptive prewhitening

作　　者：陈鹏[1] 侯朝焕[1] 马晓川[1] 梁亦慧[2] 

机构地区：[1]中国科学院声学研究所,北京100080 [2]中国船舶工业集团公司船舶系统工程部,北京100036

出　　处：《系统工程与电子技术》2007年第11期1819-1822,共4页Systems Engineering and Electronics

基　　金：国家自然科学基金资助课题(60472101)

摘　　要：有色混响噪声背景以及水下动目标径向速度造成的回波和样本失配导致匹配滤波器对于线性调频LFM(linear frequency modulation)回波检测性能下降。基于自适应预白化处理的广义似然比GLRT(generalized likelihood ratio test)方法利用混响噪声背景的自回归AR(autoregressive)模型构建白化滤波器来抑制混响噪声,但回波和混响噪声的混叠会造成AR模型偏差。结合匹配滤波的回波定位特性和基于自适应预白化处理GLRT方法的混响噪声背景抑制特性,提出结合这两种方法的联合检测算法。仿真和实验数据测试表明联合检测算法对于水下动目标LFM回波检测性能优于单纯的零速样本匹配滤波和GLRT方法。Both the colored reverberation noise background（RNB） and the mismatch between echo and replica caused by underwater moving target（UMT）＇s radial velocity will degrade matched filter（MF）＇s detection performance for linear frequency modulation（LFM） echo.The adaptive prewhitening processing（APP）-based GLRT（generalized likelihood ratio test） method utilizes whitening filters which are constructed according to RNB＇s AR（autoregressive） model to suppress RNB,but the superposition of echo and RNB will result in the deviation of AR model.Combining MF＇s echo localization property and APP-based GLRT method＇s RNB suppression property,proposes the joint detection algorithm.Simulation and experiment data tests show the joint detection algorithm performs better than the pure MF with zero radial velocity replica and the APP-based GLRT method.

关 键 词：信息处理技术 检测 线性调频回波 匹配滤波 AR模型 GLRT 

分 类 号：TN911.23[电子电信—通信与信息系统]