Review of methods for online power-distribution reconstruction

扫码查看

原文链接

NETL
NSTL
Elsevier

外文摘要：Accurate online reconstruction of reactor core power distribution is essential for ensuring the safe and efficient operation of nuclear reactors. This paper comprehensively reviews existing online power distribution reconstruction methods, which are classified into four categories: neutron diffusion equation-based methods, interpolation-based methods, probability-based methods, and machine learning-based methods. Neutron diffusion equation-based methods rely on the quasi-exact solutions of the neutron diffusion equation, or neutron transport equation if necessary and feasible, which can accurately reflect the physical processes within the reactor. Assuming equal ratios of the reconstructed to calculated values and the detector measurements to the calculations, termed the proportional function, interpolation-based methods reconstruct it over the whole core and are easy to implement. Probability-based methods effectively address measurement uncertainties using statistical frameworks, while machine learning-based methods demonstrate robust nonlinear modeling capabilities, enabling them to adapt to complex environments and dynamic conditions. This paper systematically introduces the principles of these methods and emphasizes their advantages and disadvantages.

外文关键词：

Online power-distribution reconstructionNeutron diffusion equation-based methodsInterpolation-based methodsProbability-based methodsMachine learning-based methodsRADIAL BASIS FUNCTIONSLEAST-SQUARES METHODREACTOR POWERUNCERTAINTY QUANTIFICATIONSPATIAL INTERPOLATIONNEURAL-NETWORKSCOREMODELREGULARIZATIONDECOMPOSITION

作者：

Yang, Hao、Li, Zhuo、Chen, Shengli、Wang, Tianxiang、Wang, Hairong、Chen, Shuyi

展开 >

作者单位：

Sun Yat Sen Univ

出版年：

2025

DOI：

10.1016/j.anucene.2025.111531

Annals of nuclear energy

SCI

ISSN：0306-4549

年,卷(期)：2025.221(Oct.)

参考文献量108