摘要
针对目前历史训练数据量大、非结构化特点突出、实际利用率低的问题,提出一种基于历史数据挖掘的雷达对抗知识库设计方法。从实际应用出发,对知识库设计业务需求进行了分析;以此为基础,采用多库结构形式,引入大数据思想以及信息交互思想,提出了雷达对抗知识库构建框架,对子模块进行了具体设计,并从信息挖掘、多维快速索引以及决策链路预测三个层面,设计了知识库的关键算法。提出的方法为历史数据的有效利用提供了一种新的思路和新的方法,可为组训人员进行训练设计以及电子对抗战法研究提供一定支撑。
To solve the problems that the historic data are giant and has unstructured character and the lowavailability are obvious,a design method for radar countermeasures knowledge base based on historic data isproposed.Considering the real application,the requirement of knowledge base design is firstly analyzed.Onthis basis,using the multi-base structure,the radar countermeasures knowledge base structural frame is presentedwith the big data and information interaction measures.The sub-modules are detailed designed,and thekey algorithms of knowledge base are detailed based on the information digging,the multi-dimensional quickindexing and the decision predicting layers.The method proposed provides a novel idea and measure for the effectiveuses of the historic data,which lays the foundation for the trainers making training design and electroniccountermeasures research.
作者
孙健
马辉
刘正堂
胡振震
Sun Jian;Ma Hui;Liu Zhengtang;Hu Zhenzhen(Luoyang Electronic Equipment Testing Center of China,Luoyang 471003,Henan,China)
出处
《航天电子对抗》
2019年第3期12-15,36,共5页
Aerospace Electronic Warfare
关键词
知识库
雷达对抗
霍夫曼算法
决策树
knowledge base
radar countermeasures
Hoffman algorithm
decision tree
