Machine learning-based forecasting of fission gas release in liquid metal-cooled reactor

分类： 物理学 >> 核物理学 提交时间： 2025-04-26

Li, Mr. Junfeng Yang, Mr. Jiawei Wang, Mr. Mou Zhao, Dr. Pengcheng Wang, Mr. Da Yokoyama, Dr. Ryo WANG, Dr. KAI Zhang, Dr. Guangwei Zhou, Dr. Wenzhong Zhou, Mr. Lijun

摘要： In this paper, we propose an artificial intelligence (AI)-driven framework for predicting fission gas release (FGR) in liquid metal-cooled reactors. Three AI models are trained on a comprehensive FGR database comprising 10,065 data points generated from multi-physics simulations. The results reveal that Random Forest (RF) achieves superior prediction accuracy, with a mean error (ME) of 3.96% for the optimal parameter combination (p_lin0, a_grain0, po, and t). While RF excels in database predictions, the Transformer model demonstrates exceptional capability in capturing parameter-effect trends, such as the positive correlation between FGR and fuel linear power or time. This study underscores the potential of AI models, particularly RF and Transformer, in advancing FGR forecasting for reactor safety analysis. This work can provide a replicable methodological paradigm for the FGR prediction of nuclear reactors and reliable technical support for the safety margin assessment of liquid metal-cooled reactors.

同行评议状态:待评议

 点击量 325  下载量 108  评论 0