基于属性特征的数据关联研究  

Study of Data Association Based on Attribute Characteristic

作　　者：牛曦晨 熊家军[1] 邱刚[1] 李灵芝[1] 

机构地区：[1]空军预警学院,武汉430019

出　　处：《舰船电子对抗》2014年第4期79-83,共5页Shipboard Electronic Countermeasure

基　　金：国家自然科学基金资助项目;项目编号:61102168;军队创新基金资助课题;项目编号:X11QN106

摘　　要：针对反导预警过程中弹道导弹飞行中段空间目标密集、点迹断续、目标跟踪稳定性低等特点,利用远程预警相控阵雷达和地基相控阵雷达获得的目标多特征属性信息,将属性特征数据模糊化,运用模糊相似度进行数据关联。对比三种隶属度函数对中段目标群识别正确率的高低,选取正态分布隶属度函数,计算两雷达观测目标之间的相似度,得到相似度矩阵,并按照一定的判决规则进行目标关联匹配。仿真结果表明,属性个数越多,均值差异越大,观测方差越小,关联效果越好;属性权重的合理分配也是影响关联结果的重要因素。Aiming at the characteristics such as dense spacial targets,intermittent tracks,low stabili- ty of target tracking of ballistic missile in flight mid-course of the anti-missile early warning process,multiple characteristic attributes information of target obtained by long range early-warn- ing phased array radar and ground-based phased array radar is used to make the attribute character- istic data fuzzy, and fuzzy similarity degree is used to perform data association. This paper compares the target recognition ratio to the target group in midcourse based on three membership functions, selects normal distribution membership function, calculates the similarity between targets observed by two radar, gets the similarity degree matrix, and performs target crrelation matching according to certain sentence rules. Simulation results show： the more the number of attributes, the greater the average difference, the smaller the observation variance, the better the correlation effect. The reasonable distribution of attribute weight is also the important factor influencing the correlation results.

关 键 词：属性特征 模糊相似度 数据关联 

分 类 号：TN958.92[电子电信—信号与信息处理]