Deep Learning Model Accelerates Plasma Predictions in Nuclear Fusion by 1,000 Times

Deep learning model boosts plasma predictions in nuclear fusion by 1,000 times

by JooHyeon Heo, Ulsan National Institute of Science and Technology A research team, led by Professor Jimin Lee and Professor Eisung Yoon in the Department of Nuclear Engineering at UNIST, has unveiled a deep learning-based approach that significantly accelerates the computation of a nonlinear Fokker-Planck-Landau (FPL) collision operator for fusion plasma. The findings are published in the Journal of Computational Physics. Nuclear fusion reactors, often referred to as artificial sun, rely on maintaining a high-temperature plasma environment similar to that of the sun. In this state, matter is composed of negatively charged electrons and positively charged ions. Accurately predicting the collisions between these particles is crucial for sustaining a stable fusion reaction. The plasma state is modeled using various mathematical frameworks, one of which is the FPL equation. The FPL equation predicts collisions between charged particles, known as Coulomb collisions. Traditionally, solving this equation involved iterative methods that required extensive computational time and resources. The proposed FPL-net can solve the FPL equation in a single step, achieving results 1,000 times faster than previous methods with an error margin of just one-hundred-thousandth, demonstrating exceptional accuracy. The FPL collision operation is characterized by the conservation of key physical quantities -- density, momentum, and energy. The researchers enhanced model accuracy by incorporating functions that preserve these quantities during the AI learning process. The effectiveness of the FPL-net was validated through thermal equilibrium simulations, which highlighted that accurate thermal equilibrium cannot be achieved if errors accumulate during continuous simulations. "By utilizing deep learning on GPUs, we have reduced computation time by a factor of 1,000 compared to traditional CPU-based codes," the joint research team stated. "This advancement represents a cornerstone for digital twin technologies, enabling turbulent analysis of entire nuclear fusion reactors or replicating real Tokamaks in a virtual computing environment." A Tokamak is a specialized device designed to trap plasma. While the current study focuses on electron plasma, the researchers noted that further research is needed to extend the applications of this model to more complex plasma environments containing various impurities.

Fusion plasma prediction gets 1,000x boost with deep learning model

Scientists have developed a deep learning-based approach that boosts plasma prediction on nuclear fusion by 1,000 times. The approach accelerates the computation of a nonlinear Fokker-Planck-Landau (FPL) collision operator for fusion plasma. Researchers at the Department of Nuclear Engineering at UNIST highlighted that the plasma state is modeled using various mathematical frameworks, one of which is the FPL equation. The FPL equation predicts collisions between charged particles, known as Coulomb collisions. Traditionally, solving this equation involved iterative methods that required extensive computational time and resources. "By utilizing deep learning on GPUs, we have reduced computation time by a factor of 1,000 compared to traditional CPU-based codes," said the joint research team. "This advancement represents a cornerstone for digital twin technologies, enabling turbulent analysis of entire nuclear fusion reactors or replicating real Tokamaks in a virtual computing environment." Researchers underlined that the proposed FPL-net can solve the FPL equation in a single step, achieving results 1,000 times faster than previous methods with an error margin of just one-hundred-thousandth, demonstrating exceptional accuracy. The FPL collision operation is characterized by the conservation of key physical quantities -- density, momentum, and energy. The researchers enhanced model accuracy by incorporating functions that preserve these quantities during the AI learning process, according to a press release by UNIST. Researchers also highlighted that the effectiveness of the FPL-net was validated through thermal equilibrium simulations, which highlighted that accurate thermal equilibrium cannot be achieved if errors accumulate during continuous simulations. Published in the Journal of Computational Physics, the findings reveal that researchers addressed the acceleration of the Fokker-Planck-Landau (FPL) collision operator using deep learning techniques. The developed FPL-net, a deep learning-based nonlinear Fokker-Planck-Landau collision operator, is a fully convolutional neural network optimized for computational speed with a compact model structure. FPL-net was trained on data representing various temperature conditions of an electron plasma on a two-dimensional velocity grid, ensuring generality, according to the study. Often referred to as artificial sun, nuclear fusion reactors, relies on maintaining a high-temperature plasma environment similar to that of the sun. In this state, matter is composed of negatively charged electrons and positively charged ions. Accurately predicting the collisions between these particles is crucial for sustaining a stable fusion reaction.