近日，美国劳伦斯·利弗莫尔国家实验室Kelli D. Humbird团队报道了在国家点火装置用物理知识深度学习预测聚变点火。相关论文于2025年8月14日发表在《科学》杂志上。

研究组在美国国家点火装置（National Ignition Facility）进行了一项惯性约束氦离子实验，通过产生的氦离子能量超过驱动实验的激光能量，实现了点火。值得注意的是，实验前采用的新型生成式机器学习模型——融合辐射流体力学模拟、深度学习、实验数据与贝叶斯统计方法——以超过70%的概率成功预测本次点火实验最可能实现成功。

附：英文原文

Title: Predicting fusion ignition at the National Ignition Facility with physics-informed deep learning

Author: Brian K. Spears, Scott Brandon, Dan T. Casey, John E. Field, Jim A. Gaffney, Kelli D. Humbird, Andrea L. Kritcher, Michael K. G. Kruse, Eugene Kur, Bogdan Kustowski, S. Langer, Dave Munro, Ryan Nora, J. Luc Peterson, Dave J. Schlossberg, Paul Springer, Alex Zylstra

Issue&Volume: 2025-08-14

Abstract: An inertial confinement fusion experiment, carried out at the National Ignition Facility, has achieved ignition by generating fusion energy exceeding the laser energy that drove the experiment. Prior to the experiment, a generative machine learning model that combines radiation hydrodynamics simulations, deep learning, experimental data, and Bayesian statistics was used to predict, with a probability greater than 70%, that ignition was the most likely outcome for this shot.

DOI: adm8201

Source: https://www.science.org/doi/10.1126/science.adm8201