基于雷达宽窄带多特征信息融合的空中目标识别  

Fusion Recognition of Air Target Based on DS Evidence Theory with Wideband and Narrowband Characteristics

作　　者：孙俊[1,2] 陈亚伟[1,2] 李崇谊 谢苏道[1,2] 

机构地区：[1]中国电子科技集团公司智能感知技术重点实验室,南京210039 [2]南京电子技术研究所,南京210039

出　　处：《现代雷达》2015年第7期15-19,共5页Modern Radar

摘　　要：宽带高分辨一维距离像(HRRP)含有丰富的目标特征信息,是进行型号识别的良好特征。窄带特征所携带的目标信息虽不如宽带丰富,但其与宽带特征有很强的互补性,且有识别距离远、受目标姿态影响小等优点。因此,综合利用目标的宽、窄带特征进行识别能够提高识别系统性能。文中分析了以HRRP的功率谱和目标速度、高度为识别特征的宽窄带识别系统的特点,利用DS证据理论,实现了基于宽窄带识别结果的决策层融合目标识别;外场实测数据的实验结果表明了该方法的有效性。Many target characteristics can be extracted from radar wideband high resolution range profile（ HRRP）,which is used for target recognition. Narrowband and wideband features complement each other,although narrowband features are insufficient comparing to wideband. Narrowband has the advantages of far identification range,little pose influence and so on. The system recognition performance is improved by making the information of wideband and narrowband into a whole effectively. Characteristics of recognition system are analyzed in this paper,which take the speed,height and power spectrum of HRRP as recognition features,and then the fusion recognition result of system outputs based on DS evidence theory is got. The experiment results of real data show the validity of this method.

关 键 词：高分辨一维距离像 宽带特征 窄带特征 DS证据理论 融合识别 

分 类 号：TN957.52[电子电信—信号与信息处理]