辐射防护 ›› 2021, Vol. 41 ›› Issue (S1): 59-63.

• 核与辐射事故应急准备与对策 • 上一篇    下一篇

基于朴素贝叶斯的核电厂事故诊断方法研究

齐奔1, 梁金刚1, 张立国1, 童节娟1, 闫术2   

  1. 1.清华大学核能与新能源技术研究院,北京 100084;
    2.辽宁红沿河核电有限公司,辽宁 大连 116302
  • 收稿日期:2020-12-28 出版日期:2021-10-25 发布日期:2021-11-26
  • 通讯作者: 梁金刚。E-mail:jingang@tsinghua.edu.cn
  • 作者简介:齐奔(1977—),男,清华大学硕士研究生。E-mail:qb19@mails.tsinghua.edu.cn
  • 基金资助:
    中核集团领创科研项目“支持压水堆核电厂应急决策的风险研判智能技术研究”的资助。

Research on fault diagnosis methods for nuclear power plant based on Naive Bayes

QI Ben1, LIANG Jingang1, ZHANG Liguo1, TONG Jiejuan1, YAN Shu2   

  1. 1. Institute of Nuclear and New Energy Technology, Tsinghua University,Beijing 100084;
    2. Liaoning Hongyanhe Nuclear Power Co. Ltd., Liaoning Dalian 116302
  • Received:2020-12-28 Online:2021-10-25 Published:2021-11-26

摘要: 本文将机器学习领域的贝叶斯技术应用于核应急中的电厂状态诊断,提出了基于朴素贝叶斯分类器的核电厂事故诊断方法。利用压水堆核电厂仿真机获取事故案例数据,对朴素贝叶斯分类模型进行训练,实现了对核电厂多类事故(LOCA、SGTR、MSLB)的诊断。测试结果表明,基于朴素贝叶斯分类器的核电厂事故诊断方法在诊断精度、诊断效率、事故类型可扩展性以及程序自主化诊断上有显著优势,并且模型训练中不同事故类型先验分布对诊断结果影响较小,具有较好的适用性。

关键词: 核应急, 事故诊断, 朴素贝叶斯, 机器学习

Abstract: This paper introduces Bayesian techniques from the machine learning field into the application of power plant accident diagnosis in nuclear emergency. A new approach for plant accident diagnosis based on Naive Bayes Classifier is proposed. The PWR nuclear power plant simulator is used to obtain accident case data, and the naive Bayes classification model is trained to realize the diagnosis of multiple types of accidents (LOCA, SGTR, MSLB) in nuclear power plants. The test results show that the accident diagnosis methods based on naive Bayesian classifier have significant advantages in diagnostic accuracy, diagnostic efficiency, expandability of accident types and program autonomy diagnosis. It is found the prior distribution of different accident types in model training has little influence on the training performance, indicating the good applicability of the new approach.

Key words: nuclear emergency, accident diagnosis, Naive Bayes, machine learning

中图分类号: 

  • TL73