▪ 연구실 소개: 핵융합 AI 최적화 연구실 (서재민 교수님)

​ · 홈페이지: https://sites.google.com/view/donuts-lab

​ · 구성원: 학부 연구생과 대학원생 모집 중!

▪ ​연구 방법

Our research group utilizes cutting-edge computing technologies such as AI and GPU to model nuclear fusion plasmas in tokamaks, and performs experiments based on these results for achieving stable nuclear fusion in the future. Our lab specializes in the following areas:

(1) Tokamak simulations targeting various fusion devices such as KSTAR, DIII-D, and ITER, allowing us to understand the complex plasma behavior and optimize tokamak performance.

(2) Our simulations pave the way for actual tokamak experiments, which help us validate our models and push the boundaries of what is possible in fusion research.

(3) We utilize physics-based AI, specifically Physics-Informed Neural Networks (NN), to more accurately and efficiently simulate plasma behavior, leading to more effective and optimized tokamak designs.

(4) Our lab also focuses on AI-based prediction, control, and stabilization of fusion plasmas. With the help of AI, we can better understand and control the complex behavior of fusion plasmas, making the road to practical fusion energy more tangible than ever before.

At our lab, we are passionate about solving one of the most challenging and exciting problems in modern science & technology - the development of a clean and sustainable energy source through nuclear fusion. Join us as we push the limits of fusion research and pave the way for a brighter, more sustainable future.

▪ ​최근 연구 주제/성과/수행과제

 · Observation of a new type of self-generated current in magnetized plasmas, Nature Communications 13 (2022) 6477

 · Development of operation trajectory design algorithm for control of multiple parameters using deep reinforcement learning in KSTAR, Nuclear Fusion 62 (2022) 086049

​ · Feedforward beta control in the KSTAR tokamak by deep reinforcement learning, Nuclear Fusion 61 (2021) 106010

​ · Ion heating by nonlinear Landau damping of high-n toroidal Alfvén eigenmodes in ITER, Nuclear Fusion 61 (2021) 096022

​ · Development of integrated suite of codes and its validation on KSTAR, Nuclear Fusion 61 (2021) 096020

​ · Parametric study of linear stability of toroidal Alfvén eigenmode in JET and KSTAR, Nuclear Fusion 60 (2020) 066008