基于多源信息融合的有源雷达组网方式序贯识别方法  

Sequential Method for Netting Type Recognition of Active Radars Based on Multi-source Information Fusion

作　　者：李程[1] 王伟[1] 施龙飞[1] 王雪松[1] 

机构地区：[1]国防科学技术大学电子信息系统复杂电磁环境效应国家重点实验室,长沙410073

出　　处：《电子与信息学报》2014年第10期2456-2463,共8页Journal of Electronics & Information Technology

基　　金：国家自然科学基金(61201336;41301490)资助课题

摘　　要：为了对有源雷达组网方式进行识别,该文利用Dempster-Shapher(D-S)证据理论适于解决信息序贯融合问题的特点,提出一种基于多源信息融合的有源雷达组网方式序贯识别方法。首先对侦察获取的有源雷达和通信电台信息进行特征提取,然后结合专家系统计算特征向量与雷达组网方式的匹配度,最后通过一种改进的相对加权融合算法进行证据修正融合,实现了对4种常见有源雷达组网方式的识别。仿真结果表明了该方法的有效性。In order to identify the netting type of active radar networks, the Dempster-Shapher （D-S） evidence theory is applied to solve the problem of sequential information fusion, and a sequential method for netting type recognition of active radars is presented based on multi-source information fusion. Firstly, features are extracted from the information of active radars and communication transmitters, which is obtained by reconnaissance equipments. Then the matching degrees between feature vector and different radar netting types are calculated combining with expert system. Finally, identification of 4 ordinary netting types of active radars is implemented by evidence modification and fusion using an improved mutual weight combination algorithm. Simulation results show that the proposed method is applicable and effective.

关 键 词：雷达组网 组网方式识别 证据理论 多源信息融合 专家系统 

分 类 号：TN95[电子电信—信号与信息处理]