Newswise – A learning in reactor physics is published in the journal Nuclear Science and TechnologyResearchers from Sichuan University, Shanghai Jiao Tong University, have presented two innovative neural networks to solve long-standing challenges related to K-eigenvalue problems in neutron diffusion theory. These problems, which are fundamental to the field of nuclear engineering, are crucial for the simulation and analysis of nuclear reactors.
This study presented two pioneering neural networks, the generalized inverse power method neural network (GIPMNN) and its advanced version, the physics-constrained GIPMNN (PC-GIPMNN), to solve challenges in reactor physics. While GIPMNN uses the inverse power method to iteratively specify the lowest eigenvalue and associated eigenvector, PC-GIPMNN enhances this approach by seamlessly incorporating conservative interface terms. This advance is crucial when solving the interface challenges inherent in reactors with diverse fuel assemblies. Notably, in side-by-side performance evaluations of complex spatial geometries, PC-GIPMNN consistently outperformed its counterpart GIPMNN et al. Uniquely, this study chose a data-independent approach, focusing only on mathematical and numerical solutions, thereby eliminating potential bias.
These findings herald a new era in nuclear reactor physics, paving the way for enhanced understanding and simpler simulations. The adaptability of the introduced neural networks points to their potential application in other scientific arenas that deal with interface challenges. Essentially, the research focuses on the revolutionary promise of neural networks in reactor physics. Future efforts will undoubtedly refine these networks and investigate their effectiveness in increasingly complex scenarios.
Original source URL
National Natural Science Foundation of China (11971020)
Shanghai Natural Science Foundation (23ZR1429300)
CNNC Innovation Funds (Lingchuang Fund)
Nuclear Science and Technology (NST) reports scientific advances, technical achievements and important results in the fields of nuclear science and technology. The purpose of this periodical is to stimulate the cross-fertilization of knowledge among scientists and engineers working in the fields of nuclear research.