基于CiteSpace的核电机组故障诊断发展趋势分析  

Analysis of Development Trends in Fault Diagnosis of Nuclear Power Units Based on CiteSpace

作　　者：李蔚[1] 李翱 方兴煜 卢韩斌 林小杰 尚宪和 LI Wei;LI Ao;FANG Xingyu;LU Hanbin;LIN Xiaojie;SHANG Xianhe(College of Energy Engineering,Zhejiang University,Hangzhou 310027,China;China National Nuclear Power Operation Management Co.,Ltd.,Haiyan 314300,China)

机构地区：[1]浙江大学能源工程学院,浙江杭州310027 [2]中核核电运行管理有限公司,浙江海盐314300

出　　处：《山东电力技术》2025年第3期75-85,共11页Shandong Electric Power

摘　　要：核电作为一种清洁能源,对优化我国能源结构、保障能源安全、助力实现“双碳”目标具有重要作用,而核电机组的安全运行尤其重要。文中使用CiteSpace 6.1.6软件对2009-2024年中国知网(China National Knowledge Infrastructure,CNKI)和Web of Science(WOS)数据库的核电机组故障诊断相关研究进行量化分析。研究表明:该领域国内发文以研究院为主体,国际文献则由高校领衔;故障诊断技术从专家驱动向数据驱动转变,热点趋势归纳为在线监测/性能监测-预防性维修-机器学习/主成分分析-深度学习/卷积神经网络;核电厂的数字化进程热点为仪控系统-人因工程-智能运维-数字孪生。根据文献结果分析可知,未来人工智能算法在核电故障诊断领域的应用有助于提高故障诊断精度、增强故障可解释性。As a kind of clean energy,nuclear power plays a pivotal role in optimizing energy structure,ensuring energy security,and contributing to the achievement of the"dual carbon"goals in China.Ensuring the safe operation of nuclear power units is paramount.In this paper,CiteSpace 6.1.6 software is utilized to quantitatively analyze the related research on fault diagnosis of nuclear power units in the China National Knowledge Infrastructure(CNKI)and Web of Science(WOS)databases from 2009 to 2024.The study indicates that in this field,the main body of domestic publications is the research institute,while universities take the lead of international literature.Fault diagnosis technology has transitioned from expert-driven to data-driven approaches.The hotspot trend can be summarized as online monitoring/performance monitoring,preventive maintenance,machine learning/principal component analysis,and deep learning/convolutional neural networks.The hotspots of the digitalization process of nuclear power plant are instrument control system,human factor engineering,intelligent operation and maintenance,and digital twin.According to the analysis of literature results,future applications of artificial intelligence algorithms in the field of nuclear power fault diagnosis will contribute to improving the diagnostic accuracy and enhancing interpretability.

关 键 词：核电机组 故障诊断 核电数字化 CiteSpace软件 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]