基于多目标多特征的模糊数据关联  

Fuzzy Data Association Based on Multi-Feature and Multi-Target

作　　者：牛曦晨 熊家军[1] 丁小 罗晨星 

机构地区：[1]空军预警学院 [2]解放军95865部队

出　　处：《火力与指挥控制》2015年第10期46-48,52,共4页Fire Control & Command Control

基　　金：国家自然科学基金(61102168);军队创新基金资助项目(X11QN106)

摘　　要：在弹道导弹飞行中段多目标跟踪过程中,传统的数据关联方法只利用目标距离、方位、状态等与计算直接相关的信息,针对传统数据关联存在误跟、失跟的问题,提出一种基于多目标多特征的模糊相似度关联算法。该算法同时利用目标尺寸、形状、材料3种属性特征,应用模糊数据关联计算雷达观测目标间的相似度,利用蒙特卡罗法验证其正确率。In the midcourse of ballistic missile of multiple target tracking,since the traditional algorithm only use the features directly correlative with the target such as target distance,azimuth and states. A new fuzzy association based on multi-feature and multi-target is proposed in this paper to solve the problem with low stability of target tracking. According to the information of target attribute features obtained by radar,such as target size,shape and material. The algorithm use fuzzy data association to work out the similarity between the two targets which are observed by radar. Use Monte Carlo method to verify its accuracy.

关 键 词：多目标多特征 模糊相似度 数据关联 

分 类 号：TP975[自动化与计算机技术]