中国半导体行业协会封装分会会刊

中国电子学会电子制造与封装技术分会会刊

导航

电子与封装 ›› 2019, Vol. 19 ›› Issue (4): 028 -31. doi: 10.16257/j.cnki.1681-1070.2019.0042

• 微电子制造与可靠性 • 上一篇    下一篇

基于贝叶斯网络的DC-DC电源故障不确定性分析

贾晓晓1,高会壮2,黄姣英3   

  1. 1. 火箭军驻699厂军代室,北京 100039; 2. 北京航空航天大学可靠性与系统工程学院,北京 100191
  • 收稿日期:2018-12-28 出版日期:2019-04-20 发布日期:2020-01-16
  • 作者简介:贾晓晓(1986—),女,山东潍坊人,学士,工程师,研究方向为控制系统及电子产品研制。

Uncertainty Analysis of Fault Prediction for DC-DC Power Module Based on Bayesian Network

JIA Xiaoxiao1, GAO Huizhuang2, HUANG Jiaoying2   

  1. 1. Military Delegate Office of PLA Rocket Force at Fact.699, Beijing 100039, China; 2. School of Reliability and Systems Engineering, Beihang University, Beijing 100191, China
  • Received:2018-12-28 Online:2019-04-20 Published:2020-01-16

摘要: 为解决DC-DC电源模块故障诊断中不确定性的相关问题,研究了电源模块故障不确定性产生的原因,同时对不确定性信息类型与推理方法进行了研究。利用贝叶斯网络对DC-DC电源模块故障产生原因与故障模式进行建模描述,经过BIC与K2评分算法训练完成后,可以进行不确定性推断。在故障概率方面分析了模块中的不同故障部件对电源模块故障的影响,利用MATLAB平台根据模拟数据对不确定性模型进行了分析,验证了基于贝叶斯网络的故障预测不确定性模型的有效性。

关键词: DC-DC电源模块;故障诊断;贝叶斯网络;不确定信息;不确定推理

Abstract: In order to solve the uncertainty related problems in DC-DC power module fault diagnosis, the causes of power module fault uncertainty are studied. At the same time, the uncertainty information type and reasoning method are studied. The Bayesian network is used to describe the cause and failure mode of the DC-DC power module fault. After the BIC and K2 scoring algorithms are completed, the uncertainty can be inferred. The influence of different fault components in the module on the fault of the power module is analyzed in terms of the probability of failure. The MATLAB platform is used to analyze the uncertainty model based on the simulation data, and the effectiveness of the Bayesian network-based fault prediction uncertainty model is verified.

Key words: DC-DC power supply module; fault diagnosis; Bayesian network; uncertain information; uncertain reasoning

中图分类号: