Uncovering turbulent plasma dynamics via deep learning from partial observations

Mathews, Abhilash; Francisquez, Manaure; Hughes, Jerry W.; Hatch, David R.; Zhu, Ben; Rogers, Barrett N.

doi:10.1103/physreve.104.025205

Uncovering turbulent plasma dynamics via deep learning from partial observations

Journal Article · Fri Aug 13 00:00:00 EDT 2021 · Physical Review. E

DOI:https://doi.org/10.1103/physreve.104.025205· OSTI ID:1813020

^[1]; ^[2]; ^[3]; Hatch, David R. ^[4]; Zhu, Ben ^[5]; Rogers, Barrett N. ^[6]

MIT Plasma Science and Fusion Center, Cambridge, MA (United States); MIT Plasma Science and Fusion Center
MIT Plasma Science and Fusion Center, Cambridge, MA (United States); Princeton Plasma Physics Lab. (PPPL), Princeton, NJ (United States)
MIT Plasma Science and Fusion Center, Cambridge, MA (United States)
Univ. of Texas, Austin, TX (United States)
Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)
Dartmouth College, Hanover, NH (United States)

One of the most intensely studied aspects of magnetic confinement fusion is edge plasma turbulence which is critical to reactor performance and operation. Drift-reduced Braginskii two-fluid theory has for decades been widely applied to model boundary plasmas with varying success. Towards better understanding edge turbulence in both theory and experiment, we demonstrate that a novel multi-network physics-informed deep learning framework constrained by partial differential equations can accurately learn turbulent fields consistent with the two-fluid theory from partial observations of electron pressure which is not otherwise possible using conventional equilibrium models. Furthermore, this technique presents a novel paradigm for the advanced design of plasma diagnostics and validation of magnetized plasma turbulence theories in challenging thermonuclear environments.

View Accepted Manuscript (DOE)

Research Organization:: MIT Plasma Science and Fusion Center, Cambridge, MA (United States)

Sponsoring Organization:: USDOE Office of Science (SC), Fusion Energy Sciences (FES)

Grant/Contract Number:: SC0014264; FC02-08ER54966; FG02-04ER54742; AC52-07NA27344

OSTI ID:: 1813020

Alternate ID(s):: OSTI ID: 1815370
OSTI ID: 1822274

Journal Information:: Physical Review. E, Journal Name: Physical Review. E Journal Issue: 2 Vol. 104; ISSN 2470-0045

Publisher:: American Physical Society (APS)Copyright Statement

Country of Publication:: United States

Language:: English

References (44)

On the limited memory BFGS method for large scale optimization Liu, Dong C.; Nocedal, Jorge Mathematical Programming, Vol. 45, Issue 1-3 https://doi.org/10.1007/BF01589116	journal	August 1989
Approximation by superpositions of a sigmoidal function Cybenko, G. Mathematics of Control, Signals, and Systems, Vol. 2, Issue 4 https://doi.org/10.1007/BF02551274	journal	December 1989
On the eigenvector bias of Fourier feature networks: From regression to solving multi-scale PDEs with physics-informed neural networks Wang, Sifan; Wang, Hanwen; Perdikaris, Paris Computer Methods in Applied Mechanics and Engineering, Vol. 384 https://doi.org/10.1016/j.cma.2021.113938	journal	October 2021
GDB: A global 3D two-fluid model of plasma turbulence and transport in the tokamak edge Zhu, Ben; Francisquez, Manaure; Rogers, Barrett N. Computer Physics Communications, Vol. 232 https://doi.org/10.1016/j.cpc.2018.06.002	journal	November 2018
Conceptual design study for heat exhaust management in the ARC fusion pilot plant Kuang, A. Q.; Cao, N. M.; Creely, A. J. Fusion Engineering and Design, Vol. 137 https://doi.org/10.1016/j.fusengdes.2018.09.007	journal	December 2018
Diagnostics for plasma control – From ITER to DEMO Biel, W.; Albanese, R.; Ambrosino, R. Fusion Engineering and Design, Vol. 146 https://doi.org/10.1016/j.fusengdes.2018.12.092	journal	September 2019
DGM: A deep learning algorithm for solving partial differential equations Sirignano, Justin; Spiliopoulos, Konstantinos Journal of Computational Physics, Vol. 375 https://doi.org/10.1016/j.jcp.2018.08.029	journal	December 2018
Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations Raissi, M.; Perdikaris, P.; Karniadakis, G. E. Journal of Computational Physics, Vol. 378 https://doi.org/10.1016/j.jcp.2018.10.045	journal	February 2019
Solving differential equations using deep neural networks Michoski, Craig; Milosavljević, Miloš; Oliver, Todd Neurocomputing, Vol. 399 https://doi.org/10.1016/j.neucom.2020.02.015	journal	July 2020
Non-linear simulations of the TCV Scrape-Off Layer Nespoli, F.; Furno, I.; Halpern, F. D. Nuclear Materials and Energy, Vol. 12 https://doi.org/10.1016/j.nme.2016.10.019	journal	August 2017
Continuous observation of filaments from the confined region to the far scrape-off layer Griener, M.; Wolfrum, E.; Birkenmeier, G. Nuclear Materials and Energy, Vol. 25 https://doi.org/10.1016/j.nme.2020.100854	journal	December 2020
Divertor heat flux challenge and mitigation in SPARC Kuang, A. Q.; Ballinger, S.; Brunner, D. Journal of Plasma Physics, Vol. 86, Issue 5 https://doi.org/10.1017/S0022377820001117	journal	September 2020
Automatic Differentiation in ROOT Vassilev, Vassil; Efremov, Aleksandr; Shadura, Oksana EPJ Web of Conferences, Vol. 245 https://doi.org/10.1051/epjconf/202024502015	journal	January 2020
The character of transport caused by E×B drift turbulence Scott, B. Physics of Plasmas, Vol. 10, Issue 4 https://doi.org/10.1063/1.1545100	journal	April 2003
Drift-ordered fluid equations for field-aligned modes in low-β collisional plasma with equilibrium pressure pedestals Simakov, Andrei N.; Catto, Peter J. Physics of Plasmas, Vol. 10, Issue 12 https://doi.org/10.1063/1.1623492	journal	December 2003
Intermittent fluctuations in the Alcator C-Mod scrape-off layer Garcia, O. E.; Fritzner, S. M.; Kube, R. Physics of Plasmas, Vol. 20, Issue 5 https://doi.org/10.1063/1.4802942	journal	May 2013
Full-F gyrofluid model Madsen, Jens Physics of Plasmas, Vol. 20, Issue 7 https://doi.org/10.1063/1.4813241	journal	July 2013
20 years of research on the Alcator C-Mod tokamaka) Greenwald, M.; Bader, A.; Baek, S. Physics of Plasmas, Vol. 21, Issue 11 https://doi.org/10.1063/1.4901920	journal	November 2014
Invited Review Article: Gas puff imaging diagnostics of edge plasma turbulence in magnetic fusion devices Zweben, S. J.; Terry, J. L.; Stotler, D. P. Review of Scientific Instruments, Vol. 88, Issue 4 https://doi.org/10.1063/1.4981873	journal	April 2017
Plasma particle sources due to interactions with neutrals in a turbulent scrape-off layer of a toroidally confined plasma Thrysøe, A. S.; Løiten, M.; Madsen, J. Physics of Plasmas, Vol. 25, Issue 3 https://doi.org/10.1063/1.5019662	journal	March 2018
Global turbulence simulations of the tokamak edge region with GRILLIX Stegmeir, A.; Ross, A.; Body, T. Physics of Plasmas, Vol. 26, Issue 5 https://doi.org/10.1063/1.5089864	journal	May 2019
Simulations of divertor heat flux width using transport code with cross-field drifts under the BOUT++ framework Li, N. M.; Xu, X. Q.; Hughes, J. W. AIP Advances, Vol. 10, Issue 1 https://doi.org/10.1063/1.5126884	journal	January 2020
Collisional multispecies drift fluid model Poulsen, A.; Rasmussen, J. Juul; Wiesenberger, M. Physics of Plasmas, Vol. 27, Issue 3 https://doi.org/10.1063/1.5140522	journal	March 2020
New fluid model for the turbulent transport due to the ion temperature gradient Kim, Chang‐Bae; Horton, Wendell; Hamaguchi, Satoshi Physics of Fluids B: Plasma Physics, Vol. 5, Issue 5 https://doi.org/10.1063/1.860891	journal	May 1993
Dynamics of seeded blobs under the influence of inelastic neutral interactions Thrysøe, A. S.; Naulin, V.; Nielsen, A. H. Physics of Plasmas, Vol. 27, Issue 5 https://doi.org/10.1063/5.0003262	journal	May 2020
Fluid and gyrokinetic turbulence in open field-line, helical plasmas Francisquez, M.; Bernard, T. N.; Zhu, B. Physics of Plasmas, Vol. 27, Issue 8 https://doi.org/10.1063/5.0005333	journal	August 2020
Constructing a new predictive scaling formula for ITER's divertor heat-load width informed by a simulation-anchored machine learning Chang, C. S.; Ku, S.; Hager, R. Physics of Plasmas, Vol. 28, Issue 2 https://doi.org/10.1063/5.0027637	journal	February 2021
Plasma-material interactions in current tokamaks and their implications for next step fusion reactors Federici, G.; Skinner, C. H.; Brooks, J. N. Nuclear Fusion, Vol. 41, Issue 12 https://doi.org/10.1088/0029-5515/41/12/218	journal	December 2001
Evidence for electromagnetic fluid drift turbulence controlling the edge plasma state in the Alcator C-Mod tokamak LaBombard, B.; Hughes, J. W.; Mossessian, D. Nuclear Fusion, Vol. 45, Issue 12 https://doi.org/10.1088/0029-5515/45/12/022	journal	November 2005
Evidence for the neoclassical nature of the radial electric field in the edge transport barrier of ASDEX Upgrade Viezzer, E.; Pütterich, T.; Angioni, C. Nuclear Fusion, Vol. 54, Issue 1 https://doi.org/10.1088/0029-5515/54/1/012003	journal	December 2013
A first-principles self-consistent model of plasma turbulence and kinetic neutral dynamics in the tokamak scrape-off layer Wersal, C.; Ricci, P. Nuclear Fusion, Vol. 55, Issue 12 https://doi.org/10.1088/0029-5515/55/12/123014	journal	November 2015
Kinetic calculation of neoclassical transport including self-consistent electron and impurity dynamics Belli, E. A.; Candy, J. Plasma Physics and Controlled Fusion, Vol. 50, Issue 9 https://doi.org/10.1088/0741-3335/50/9/095010	journal	July 2008
Simulation of plasma turbulence in scrape-off layer conditions: the GBS code, simulation results and code validation Ricci, P.; Halpern, F. D.; Jolliet, S. Plasma Physics and Controlled Fusion, Vol. 54, Issue 12 https://doi.org/10.1088/0741-3335/54/12/124047	journal	November 2012
On the validity of drift-reduced fluid models for tokamak plasma simulation Leddy, Jarrod; Dudson, Ben; Romanelli, Michele Plasma Physics and Controlled Fusion, Vol. 57, Issue 12 https://doi.org/10.1088/0741-3335/57/12/125016	journal	November 2015
Development and simulation of multi-diagnostic Bayesian analysis for 2D inference of divertor plasma characteristics Bowman, C.; Harrison, J. R.; Lipschultz, B. Plasma Physics and Controlled Fusion, Vol. 62, Issue 4 https://doi.org/10.1088/1361-6587/ab759b	journal	February 2020
Electric field and turbulence in global Braginskii simulations across the ASDEX Upgrade edge and scrape-off layer Zholobenko, W.; Body, T.; Manz, P. Plasma Physics and Controlled Fusion, Vol. 63, Issue 3 https://doi.org/10.1088/1361-6587/abd97e	journal	February 2021
Global 3D Braginskii simulations of the tokamak edge region of IWL discharges Francisquez, M.; Zhu, B.; Rogers, B. N. Nuclear Fusion, Vol. 57, Issue 11 https://doi.org/10.1088/1741-4326/aa7f23	journal	August 2017
Padé-based arbitrary wavelength polarization closures for full-F gyro-kinetic and -fluid models Held, M.; Wiesenberger, M.; Kendl, A. Nuclear Fusion, Vol. 60, Issue 6 https://doi.org/10.1088/1741-4326/ab7a68	journal	May 2020
UKAEA capabilities to address the challenges on the path to delivering fusion power Chapman, I. T.; Morris, A. W. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, Vol. 377, Issue 2141 https://doi.org/10.1098/rsta.2017.0436	journal	February 2019
First Evidence of Local E × B Drift in the Divertor Influencing the Structure and Stability of Confined Plasma near the Edge of Fusion Devices Wang, H. Q.; Guo, H. Y.; Xu, G. S. Physical Review Letters, Vol. 124, Issue 19 https://doi.org/10.1103/PhysRevLett.124.195002	journal	May 2020
Curvature of Radial Electric Field Aggravates Edge Magnetohydrodynamics Mode in Toroidally Confined Plasmas Zhang, Y.; Guo, Z. B.; Diamond, P. H. Physical Review Letters, Vol. 125, Issue 25 https://doi.org/10.1103/PhysRevLett.125.255003	journal	December 2020
Progress in anomalous transport research in toroidal magnetic confinement devices Carreras, B. A. IEEE Transactions on Plasma Science, Vol. 25, Issue 6 https://doi.org/10.1109/27.650902	journal	January 1997
Artificial neural networks for solving ordinary and partial differential equations Lagaris, I. E.; Likas, A.; Fotiadis, D. I. IEEE Transactions on Neural Networks, Vol. 9, Issue 5 https://doi.org/10.1109/72.712178	journal	January 1998
Quantifying experimental edge plasma evolution via multidimensional adaptive Gaussian process regression Abhilash Mathews, Jerry W. Hughes Harvard Dataverse https://doi.org/10.7910/dvn/yngfyk	dataset	January 2021

Cited By (1)

Quantifying Experimental Edge Plasma Evolution Via Multidimensional Adaptive Gaussian Process Regression Mathews, Abhilash; Hughes, Jerry W. IEEE Transactions on Plasma Science, Vol. 49, Issue 12 https://doi.org/10.1109/tps.2021.3123046	journal	December 2021

Similar Records

Uncovering turbulent plasma dynamics via deep learning from partial observations

Dataset · Mon May 24 00:00:00 EDT 2021 · OSTI ID:1882600

Related Subjects

70 PLASMA PHYSICS AND FUSION TECHNOLOGY
Artificial neural networks
Magnetic confinement fusion
Plasma turbulence

Uncovering turbulent plasma dynamics via deep learning from partial observations

Citation Formats

References (44)

Cited By (1)

Similar Records

Related Subjects