计算机技术与发展2017，Vol.27Issue(9)：17-21,5.DOI:10.3969/j.issn.1673-629X.2017.09.004
基于模糊变结构动态贝叶斯网的目标识别方法
A Target Identification Method of Dynamic Bayesian Network with Fuzzy Variable Structure
摘要
Abstract
By researching and analyzing the characteristics of the source information and the basic process of target identification,on the basis of traditional static Bayesian network,a method of target identification based on fuzzy variable structure dynamic Bayesian network is proposed. It constructs the fuzzy variable structure dynamic Bayesian network and proposes a statistical method based on sample infor-mation and a learning method of sample-free Bayesian network parameters for implementation of target identification according to net-work inference and application of traditional hard decision. The dynamic decision has been performed based on the soft decision principles and the network parameters' update online is finished based on liner weighting theory. Compared with traditional static Bayesian network for target identification, it has solved the issues such as the sequential relationship of evidence at different time and the networks inference of constant random variables. Meanwhile it has not only improved the confidence coefficient of target identification but also shortened the identification convergence period and effectively resolved error identification problem caused by error or ambiguity association. In addi-tion,the problem of network parameters unchanged has been solved and the network parameters' update online has also been completed.关键词
数据融合/目标识别/贝叶斯网络/结构学习/参数学习Key words
data fusion/target identification/Bayesian network/structure learning/parameter learning分类
信息技术与安全科学引用本文复制引用
高晓利,李捷..基于模糊变结构动态贝叶斯网的目标识别方法[J].计算机技术与发展,2017,27(9):17-21,5.基金项目
国防预先研究项目(12100201) （12100201）
