基于贝叶斯网络的电子系统电磁脉冲易损性信息融合方法  

Vulnerability Information Fusion Approach of Multi-Level HierarchicalSystems to EMP Based on Bayesian Networks

作　　者：刘钰 闫凯 杜配冰 韩峰 LIU Yu;YAN Kai;DU Pei-bing;HAN Feng(Northwest Institute of Nuclear Technology,Xi’an 710024,China)

机构地区：[1]西北核技术研究院,西安710024

出　　处：《现代应用物理》2020年第1期86-96,共11页Modern Applied Physics

基　　金：国家自然科学基金重点基金资助项目(61231003);国家自然科学基金青年基金资助项目(61201090)。

摘　　要：在电子系统电磁脉冲易损性评估问题中,电子系统各层元素的易损性信息具有显著的不均衡性,为了降低系统电磁脉冲易损性评估计算过程中的不确定度,本文基于分层贝叶斯网络模型及贝叶斯信息融合算法,建立了电子系统电磁脉冲易损性评估模型及易损性信息融合算法,并以某星用模数转换系统的电磁脉冲易损性评估问题为例,应用本文提出的信息融合算法,给出了系统后验概率分布及后验均值估计,验证了本文计算的有效性。The vulnerability information available at different system levels has a significant imbalance in the vulnerability assessment of electronic systems to electromagnetic pulse(EMP).To reduce the uncertainty in the calculation of electromagnetic pulse vulnerability assessment,this paper presents an electromagnetic pulse vulnerability assessment model and a vulnerability information fusion algorithm in an electronic system,and takes the electromagnetic pulse vulnerability assessment of an analog-to-digital conversion system as an example based on the hierarchical Bayesian networks and Bayesian information fusion approach.By using the information fusion algorithm proposed in this paper,the posterior probability distribution and posterior mean estimation of the system are given,and the validity of the calculation is verified.

关 键 词：电磁脉冲易损性 信息融合算法 分层贝叶斯网络 

分 类 号：TN07[电子电信—物理电子学] N945.17[自然科学总论—系统科学]