Machine learning for analysis of real nuclear plant data in the frequency domain
Annals of Nuclear Energy(2022)
摘要
•Machine learning is used to effectively detect anomalies in nuclear power plants.•Deep neural networks are trained using simulated data in the frequency domain.•Training data are generated by the CORE SIM+ neutronic modelling tool.•Domain adaptation methods are used to align simulated and real nuclear plant data.•Unsupervised and self-supervised methods detect anomalies’ types and locations.
更多查看译文
关键词
Neutron noise,Machine learning,Domain adaptation,Unsupervised learning,Clustering,Self-supervised learning,Core diagnostics,Core monitoring,Simulated data,Actual plant data
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要