Implementation of AI/Deep Learning Disruption Predictor into a Plasma Control System
William Tang, Ge Dong, Jayson Barr, Keith Erickson, Rory Conlin, M. Dan Boyer, Julian Kates-Harbeck, Kyle Felker, Cristina Rea, Nikolas C. Logan, Alexey Svyatko… 展开作者
William Tang, Ge Dong, Jayson Barr, Keith Erickson, Rory Conlin, M. Dan Boyer, Julian Kates-Harbeck, Kyle Felker, Cristina Rea, Nikolas C. Logan, Alexey Svyatkovskiy, Eliot Feibush, Joseph Abbatte, Mitchell Clement, Brian Grierson, Raffi Nazikian, Zhihong Lin, David Eldon, Auna Moser, Mikhail Maslov
本文报道了基于融合循环神经网络(Fusion Recurrent Neural Network,FRNN)的前沿深度学习破裂预测模型的最新进展,FRNN最初于2019年发表在《自然》期刊上。具体而言,该预测器如今不仅能输出破裂评分(作为即将发生破裂的概率指标),还能实时生成敏感度评分,用以指示即将发生破裂的潜在原因。这为深度学习模型增添了宝贵的物理可解释性,且由于该模型已完全集成到现代等离子体控制系统(Plasma Control System,PCS)中,能够为控制执行器提供有用的指导。这一进展是从现代深度学习破裂预测迈向实时控制的重要一步,并带来了适用于未来燃烧等离子体ITER系统的新型AI赋能功能。我们的分析使用了来自JET和DIII-D的大量数据,这些数据已在早期《自然》期刊论文中经过验证。除了预测放电何时会发生破裂外,本文还通过开展敏感度研究,探究了破裂发生的原因。据此,FRNN得到了扩展,可使用更多通道的信息,包括测得的DIII-D信号,例如:(i)n1rms信号,该信号与具有有限频率的n=1模相关,包括新经典撕裂模和锯齿动力学;(ii)指示等离子体杂质含量的 bolometer(辐射热计)数据;(iii)q-min,即与扭结模关键物理过程相关的安全因子最小值。新增的信号通道和可解释性功能,拓展了深度学习FRNN软件的能力,使其能够提供破裂子类别相关信息,并为等离子体控制系统中的执行器提供更精准、更直接的指导。
英文摘要:
This paper reports on advances to the state-of-the-art deep-learning disruption prediction models based on the Fusion Recurrent Neural Network (FRNN) originally introduced a 2019 Nature publication. In particular, the predictor now features not only the disruption score, as an indicator of the probability of an imminent disruption, but also a sensitivity score in real-time to indicate the underlying reasons for the imminent disruption. This adds valuable physics-interpretability for the deep-learning model and can provide helpful guidance for control actuators now that it is fully implemented into a modern Plasma Control System (PCS). The advance is a significant step forward in moving from modern deep-learning disruption prediction to real-time control and brings novel AI-enabled capabilities relevant for application to the future burning plasma ITER system. Our analyses use large amounts of data from JET and DIII-D vetted in the earlier NATURE publication. In addition to when a shot is predicted to disrupt, this paper addresses reasons why by carrying out sensitivity studies. FRNN is accordingly extended to use many more channels of information, including measured DIII-D signals such as (i) the n1rms signal that is correlated with the n =1 modes with finite frequency, including neoclassical tearing mode and sawtooth dynamics, (ii) the bolometer data indicative of plasma impurity content, and (iii) q-min, the minimum value of the safety factor relevant to the key physics of kink modes. The additional channels and interpretability features expand the ability of the deep learning FRNN software to provide information about disruption subcategories as well as more precise and direct guidance for the actuators in a plasma control system.
Scenario adaptive disruption prediction study for next generation burning-plasma tokamaks
Plasma Science and Fusion Center, Massachusetts Institute of Technology(麻省理工学院等离子体科学与聚变中心);Institute of Plasma Physics, Chinese Academy of Sciences(中国科学院等离子体物理研究所);General Atomics(通用原子公司)
Plasma State Monitoring and Disruption Characterization using Multimodal VAEs
1 École Polytechnique Fédérale de Lausanne (EPFL), Swiss Plasma Center (SPC), CH-1015 Lausanne, Switzerland;2 Eindhoven University of Technology (TU/e), Mathematics;Computer Science, NL-5600MB Eindhoven, The Netherlands;3 Swiss Data Science Center (SDSC), ETH Zürich \& EPFL, CH-8092 Zürich \& CH-1015 Lausanne, Switzerland
Evolution of SPI-induced disruptions in ASDEX Upgrade
Max Planck Institute for Plasma Physics(马克斯·普朗克等离子体物理研究所);Institute of Plasma Physics of the CAS(捷克科学院等离子体物理研究所);Institute for Applied Physics(应用物理研究所)
Collisional-radiative data for tokamak disruption mitigation modeling
Los Alamos National Laboratory(洛斯阿拉莫斯国家实验室);Curtin University(科廷大学);Korea Institute of Fusion Energy(韩国聚变能源研究所);Queensland Quantum and Advanced Technologies Research Institute, Griffith University(格里菲大学昆士兰量子与先进技术研究学院);School of Environment and Science, Griffith University(格里菲大学环境与科学学院)
Active Disruption Avoidance and Trajectory Design for Tokamak Ramp-downs with Neural Differential Equations and Reinforcement Learning
Plasma Science and Fusion Center, Massachusetts Institute of Technology(麻省理工学院等离子体科学与聚变中心);Laboratory for Information and Decision Systems, Massachusetts Institute of Technology(麻省理工学院信息与决策系统实验室);OpenStar Technologies LLC.(OpenStar科技有限责任公司)
Resistive wall mode induced disruptions in an advanced tokamak
State Key Laboratory of Advanced Electromagnetic Technology(先进电磁技术国家重点实验室);International Joint Research Laboratory of Magnetic Confinement Fusion and Plasma Physics(磁约束聚变与等离子体物理国际联合研究中心)
Non-thermal electron cyclotron emission during runaway plateau in tokamak disruptions from a highly anisotropic dielectric tensor
Seoul National University(首尔大学);Nuclear Research Institute for Future Technology and Policy, Seoul National University(首尔大学未来技术与政策核研究所);POSTECH(浦项工科大学)