Within integrated tokamak plasma modeling, turbulent transport codes are typically the computational bottleneck limiting their routine use outside of post-discharge analysis. Neural network (NN) surrogates have been used to accelerate these calculations while retaining the desired accuracy of the physics-based models. This paper extends a previous NN model, known as QLKNN-hyper-10D, by incorporating the impact of impurities, plasma rotation, and magnetic equilibrium effects. This is achieved by adding a light impurity fractional density (nimp,light/ne) and its normalized gradient, the normalized pressure gradient (α), the toroidal Mach number (Mtor), and the normalized toroidal flow velocity gradient. The input space was sampled based on experimental data from the JET tokamak to avoid the curse of dimensionality. The resulting networks, named QLKNN-jetexp-15D, show good agreement with the original QuaLiKiz model, both by comparing individual transport quantity predictions and by comparing its impact within the integrated model, JINTRAC. The profile-averaged RMS of the integrated modeling simulations is <10% for each of the five scenarios tested. This is non-trivial given the potential numerical instabilities present within the highly nonlinear system of equations governing plasma transport, especially considering the novel addition of momentum flux predictions to the model proposed here. An evaluation of all 25 NN output quantities at one radial location takes ∼0.1 ms, 104 times faster than the original QuaLiKiz model. Within the JINTRAC integrated modeling tests performed in this study, using QLKNN-jetexp-15D resulted in a speed increase of only 60–100 as other physics modules outside of turbulent transport become the bottleneck.

1.
J.
Kates-Harbeck
,
A.
Svyatkovskiy
, and
W.
Tang
, “
Predicting disruptive instabilities in controlled fusion plasmas through deep learning
,”
Nature
568
,
526
(
2019
).
2.
C.
Rea
,
R. S.
Granetz
,
K.
Montes
,
R. A.
Tinguely
,
N.
Eidietis
,
J. M.
Hanson
, and
B.
Sammuli
, “
Disruption prediction investigations using machine learning tools on DIII-D and Alcator C-Mod
,”
Plasma Phys. Controlled Fusion
60
,
084004
(
2018
).
3.
D.
Böckenhoff
,
M.
Blatzheim
,
H.
Hölbe
,
H.
Niemann
,
F.
Pisano
,
R.
Labahn
, and
T.
Pedersen
, “
Reconstruction of magnetic configurations in W7-X using artificial neural networks
,”
Nucl. Fusion
58
,
056009
(
2018
).
4.
V.
Škvára
,
V.
Šmídl
, and
J.
Urban
, “
Robust sparse linear regression for tokamak plasma boundary estimation using variational Bayes
,”
J. Phys.: Conf. Ser.
1047
,
012015
(
2018
).
5.
L.
Hesslow
,
L.
Unnerfelt
,
O.
Vallhagen
,
O.
Embréus
,
M.
Hoppe
,
G.
Papp
, and
T.
Fulöp
, “
Evaluation of the Dreicer runaway generation rate in the presence of high-Z impurities using a neural network
,”
J. Plasma Phys.
85
,
475850601
(
2019
).
6.
J.
Schmidhuber
, “
Deep learning in neural networks: An overview
,”
Neural Netw.
61
,
85
(
2015
).
7.
J.
Citrin
,
S.
Breton
,
F.
Felici
,
F.
Imbeaux
,
T.
Aniel
,
J.
Artaud
,
B.
Baiocchi
,
C.
Bourdelle
,
Y.
Camenen
, and
J.
Garcia
, “
Real-time capable First principle based modelling of tokamak turbulent transport
,”
Nucl. Fusion
55
,
092001
(
2015
).
8.
O.
Meneghini
,
S.
Smith
,
P.
Snyder
,
G.
Staebler
,
J.
Candy
,
E.
Belli
,
L.
Lao
,
M.
Kostuk
,
T.
Luce
,
T.
Luda
,
J.
Park
, and
F.
Poli
, “
Self-consistent corepedestal transport simulations with neural network accelerated models
,”
Nucl. Fusion
57
,
086034
(
2017
).
9.
J.
Garcia
,
C.
Challis
,
D.
Gallart
,
L.
Garzotti
,
T.
Görler
,
D.
King
, and
M.
Mantsinen
, “
Challenges in the extrapolation from DD to DT plasmas: Experimental analysis and theory based predictions for JET-DT
,”
Plasma Phys. Controlled Fusion
59
,
014023
(
2017
).
10.
F.
Casson
,
H.
Patten
,
C.
Bourdelle
,
S.
Breton
,
J.
Citrin
,
F.
Koechl
,
M.
Sertoli
,
C.
Angioni
,
Y.
Baranov
,
R.
Bilato
,
E.
Belli
,
C.
Challis
,
G.
Corrigan
,
A.
Czarnecka
,
O.
Ficker
,
L.
Frassinetti
,
L.
Garzotti
,
M.
Goniche
,
J.
Graves
,
T.
Johnson
,
K.
Kirov
,
P.
Knight
,
E.
Lerche
,
M.
Mantsinen
,
J.
Mylnar
, and
M.
Valisa
, “
Predictive multi-channel uxdriven modelling to optimise ICRH tungsten control and fusion performance in JET
,”
Nucl. Fusion
60
,
066029
(
2020
).
11.
S.
Breton
,
F.
Casson
,
C.
Bourdelle
,
J.
Citrin
,
Y.
Baranov
,
Y.
Camenen
,
C.
Challis
,
G.
Corrigan
,
J.
Garcia
,
L.
Garzotti
,
S.
Henderson
,
F.
Koechl
,
E.
Militello-Asp
,
M.
OMullane
,
T.
Pütterich
,
M.
Sertoli
, and
M.
Valisa
, “
First principle integrated modeling of multi-channel transport including tungsten in JET
,”
Nucl. Fusion
58
,
096003
(
2018
).
12.
O.
Linder
,
J.
Citrin
,
G.
Hogeweij
,
C.
Angioni
,
C.
Bourdelle
,
F.
Casson
,
E.
Fable
,
A.
Ho
,
F.
Koechl
,
M.
Sertoli
,
EUROfusion MST1 Team, and ASDEX Upgrade team
, “
Flux-driven integrated modelling of main ion pressure and trace tungsten transport in ASDEX Upgrade
,”
Nucl. Fusion
59
,
016003
(
2019
).
13.
M.
Romanelli
,
G.
Corrigan
,
V.
Parail
,
S.
Wiesen
,
R.
Ambrosino
,
P.
Da Silva Aresta Belo
,
L.
Garzotti
,
P.
Harting
,
F.
Köchl
,
T.
Koskela
,
L.
Lauro-Taroni
,
C.
Marchetto
,
M.
Mattei
,
E.
Militello-Asp
,
M.
Nave
,
S.
Pamela
,
A.
Salmi
,
P.
Strand
, and
G.
Szepesi
, “
JINTRAC: A system of codes for integrated simulation of tokamak scenarios
,”
Plasma Fusion Res.
9
,
3403023
(
2014
).
14.
F.
Jenko
,
W.
Dorland
,
M.
Kotschenreuther
, and
B. N.
Rogers
, “
Electron temperature gradient driven turbulence
,”
Phys. Plasmas
7
,
1904
1910
(
2000
).
15.
A. G.
Peeters
,
Y.
Camenen
,
F. J.
Casson
,
W. A.
Hornsby
,
A. P.
Snodin
,
D.
Strintzi
, and
G.
Szepesi
, “
The nonlinear gyro-kinetic flux tube code GKW
,”
Comput. Phys. Commun.
180
,
2650
2672
(
2009
).
16.
C.
Bourdelle
,
J.
Citrin
,
B.
Baiocchi
,
A.
Casati
,
P.
Cottier
,
X.
Garbet
,
F.
Imbeaux
, and
J. Contributors,
Core turbulent transport in tokamak plasmas: Bridging theory and experiment with Qua-LiKiz
,”
Plasma Phys. Controlled Fusion
58
,
014036
(
2016
).
17.
K. L.
van de Plassche
,
J.
Citrin
,
C.
Bourdelle
,
Y.
Camenen
,
F. J.
Casson
,
V. I.
Dagnelie
,
F.
Felici
,
A.
Ho
, and
S.
Van Mulders
, “
Fast modeling of turbulent transport in fusion plasmas using neural networks
,”
Phys. Plasmas
27
,
022310
(
2020
).
18.
P.
Ennever
,
M.
Porkolab
,
J.
Candy
,
G.
Staebler
,
M.
Reinke
,
J.
Rice
,
J.
Rost
,
D.
Ernst
,
C.
Fiore
,
J.
Hughes
, and
J.
Terry
, “
The effects of dilution on turbulence and transport in C-Mod ohmic plasmas and comparisons with gyrokinetic simulations
,”
Phys. Plasmas
22
,
072507
(
2015
).
19.
Y.
Camenen
,
F. J.
Casson
,
P.
Manas
, and
A. G.
Peeters
, “
Interplay between toroidal rotation and flow shear in turbulence stabilisation
,”
Phys. Plasmas
23
,
022507
(
2016
).
20.
J.
Citrin
,
C.
Bourdelle
,
F. J.
Casson
,
C.
Angioni
,
N.
Bonanomi
,
Y.
Camenen
,
X.
Garbet
,
L.
Garzotti
,
T.
Görler
,
O.
Gürcan
,
F.
Koechl
,
F.
Imbeaux
,
O.
Linder
,
K.
van de Plassche
,
P.
Strand
,
G.
Szepesi
, and
J.
Contributors
, “
Tractable flux-driven temperature, density, and rotation profile evolution with the quasilinear gyrokinetic transport model Qua-LiKiz
,”
Plasma Phys. Controlled Fusion
59
,
124005
(
2017
).
21.
O.
Meneghini
,
G.
Snoep
,
B. C.
Lyons
,
J.
McClenaghan
,
C. S.
Imai
,
B.
Grierson
,
S. P.
Smith
,
G. M.
Staebler
,
P. B.
Snyder
,
J.
Candy
,
E.
Belli
,
L.
Lao
,
J. M.
Park
,
J.
Citrin
,
T. L.
Cordemiglia
,
A.
Tema
, and
S.
Mordijck
, “
Neural-network accelerated coupled core-pedestal simulations with self-consistent transport of impurities and compatible with ITER IMAS
,”
Nucl. Fusion
61
(
2
),
026006
(
2020
).
22.
A.
Casati
,
C.
Bourdelle
,
X.
Garbet
,
F.
Imbeaux
,
J.
Candy
,
F.
Clairet
,
G.
Dif-Pradalier
,
G.
Falchetto
,
T.
Gerbaud
,
V.
Grandgirard
,
ö
Gürcan
,
P.
Hennequin
,
J.
Kinsey
,
M.
Ottaviani
,
R.
Sabot
,
Y.
Sarazin
,
L.
Vermare
, and
R.
Waltz
, “
Validating a quasi-linear transport model versus nonlinear simulations
,”
Nucl. Fusion
49
,
085012
(
2009
).
23.
J.
Citrin
,
C.
Bourdelle
,
P.
Cottier
,
D. F.
Escande
,
ö. D.
Gürcan
,
D. R.
Hatch
,
G. M. D.
Hogeweij
,
F.
Jenko
, and
M. J.
Pueschel
, “
Quasilinear transport modelling at low magnetic shear
,”
Phys. Plasmas
19
,
062305
(
2012
).
24.
See https://gitlab.com/qualikiz-group/QuaLiKiz/-/wikis/Input-and-output-variables for Input and output variables; last accessed 30 October
2020
.
25.
T.
Hahm
,
D.
Na
,
J.
Lee
,
J.
Park
,
Y.-S.
Na
,
S.
Kim
,
W.
Ko
,
P.
Diamond
,
H.
Jhang
, and
Y.
Jeon
, “
ExB shear suppression of turbulence in diverted H-mode plasmas: Role of edge magnetic shear
,”
Nucl. Fusion
53
,
093005
(
2013
).
26.
S.
Weinzierl
, “
Introduction to Monte Carlo methods
,” in
Explorations in Monte Carlo Methods
(
Springer
,
2009
), pp. 243.
27.
F.
James
, “
Monte Carlo theory and practice
,”
Rep. Prog. Phys.
43
,
1145
1189
(
1980
).
28.
M.
Mckay
,
R.
Beckman
, and
W.
Conover
, “
A comparison of three methods for selecting vales of input variables in the analysis of output from a computer code
,”
Technometrics
21
,
239
245
(
1979
).
29.
M.
Ester
,
H.-P.
Kriegel
,
J.
Sander
, and
X.
Xu
, “
A density-based algorithm for discovering clusters in large spatial databases with noise
,” in
Proceedings of the Second International Conference on Knowledge Discovery and Data Mining
(
1996
), pp.
226
231
.
30.
J.
MacQueen
, “
Some methods for classification and analysis of multivariate observations
,” in
Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability
(
1967
), Vol.
1
, pp.
281
297
.
31.
A.
Ho
,
J.
Citrin
,
F.
Auriemma
,
C.
Bourdelle
,
F. J.
Casson
,
H.-T.
Kim
,
P.
Manas
,
G.
Szepesi
, and
H.
Weisen
, “
Application of Gaussian process regression to plasma turbulent transport model validation via integrated modelling
,”
Nucl. Fusion
59
,
056007
(
2019
).
32.
C.
Rasmussen
and
C.
Williams
,
Gaussian Processes for Machine Learning
(
MIT Press
,
Cambridge, MA
,
2006
).
33.
S. C.
Guo
and
F.
Romanelli
, “
The linear threshold of the ion-temperature-gradient-driven mode
,”
Phys. Fluids B
5
,
520
533
(
1993
).
34.
P.
Sirén
,
J.
Varje
,
H.
Weisen
, and
L.
Giacomelli
, “
Role of JETPEAK database in validation of synthetic neutron camera diagnostics and ASCOTAFSI fast particle and fusion product calculation chain in JET
,”
J. Instrum.
14
,
C11013
(
2019
).
35.
E.
Cureton
, “
Unbiased estimation of the standard deviation
,”
Am. Stat.
22
,
22
(
1968
).
36.
A.
Ho
and
K. L.
van de Plassche
, https://gitlab.com/qualikiz-group/qlknn-jetexp.git for QLKNN-jetexp; last accessed 28 January
2021
.
37.
S.
Haykin
,
Neural Networks: A Comprehensive Foundation
, 2nd ed. (
Prentice Hall PTR
,
1998
).
38.
Y.
Lecun
,
L.
Bottou
,
Y.
Bengio
, and
P.
Haffner
, “
Gradient-based learning applied to document recognition
,”
Proc. IEEE
86
,
2278
2324
(
1998
).
39.
I.
Goodfellow
,
Y.
Bengio
, and
A.
Courville
,
Deep Learning
(
MIT Press
,
2016
).
40.
K.
Jarrett
,
K.
Kavukcuoglu
,
M.
Ranzato
, and
Y.
LeCun
, “
What is the best multi-stage architecture for object recognition?
,” in
2009 IEEE 12th International Conference on Computer Vision
(
2009
), pp.
2146
2153
.
41.
T.
Masters
,
Practical Neural Network Recipes in c++
(
Academic Press Professional, Inc.
,
1993
).
42.
D.
Stathakis
, “
How many hidden layers and nodes?
,”
Int. J. Remote Sens.
30
,
2133
2147
(
2009
).
43.
D. E.
Rumelhart
,
G. E.
Hinton
, and
R. J.
Williams
, “
Learning representations by back-propagating errors
,”
Nature
323
,
533
536
(
1986
).
44.
M.
Boyer
,
S.
Kaye
, and
K.
Erickson
, “
Real-time capable modeling of neutral beam injection on NSTX-u using neural networks
,”
Nucl. Fusion
59
,
056008
(
2019
).
45.
M.
Marin
,
J.
Citrin
,
C.
Bourdelle
,
Y.
Camenen
,
F.
Casson
,
A.
Ho
,
F.
Koechl
, and
M.
Maslov
, “
Firstprinciples-based multiple-isotope particle transport modelling at JET
,”
Nucl. Fusion
60
,
046007
(
2020
).
46.
G.
Cenacchi
and
A.
Taroni
, “
JETTO: A free-boundary plasma transport code
,” Technical Report No. ENEA-RT-TIB–88-5, JETIR,
1988
.
47.
M.
Nave
,
S.
Ali-Arshad
,
B.
Alper
,
B.
Balet
,
H. D.
Blank
,
D.
Borba
,
C.
Challis
,
M. V.
Hellermann
,
T.
Hender
,
G.
Huysmans
,
W.
Kerner
,
G.
Kramer
,
F.
Porcelli
,
J.
O'Rourke
,
L.
Porte
,
G.
Sadler
,
P.
Smeulders
,
A.
Sips
,
P.
Stubberfield
,
D.
Stork
,
R.
Reichle
,
J.
Wessom
, and
W.
Zwingmann
, “
MHD activity in JET hot ion H mode discharges
,”
Nucl. Fusion
35
,
409
429
(
1995
).
48.
A.
Bécoulet
,
L.-G.
Eriksson
,
Y.
Baranov
,
D.
Borba
,
C.
Challis
,
G.
Conway
,
V.
Fuchs
,
C.
Gormezano
,
C.
Gowers
,
N.
Hawkes
,
T.
Hender
,
G.
Huysmans
,
E.
Joffrin
,
X.
Litaudon
,
P.
Lomas
,
A.
Maas
,
M.
Mayoral
,
V.
Parail
,
F.
Rimini
,
F.
Rochard
,
Y.
Sarazin
,
A.
Sips
,
F.
Söldner
,
K.-D.
Zastrow
, and
W.
Zwingman
, “
Performance and control of optimized shear discharges in JET
,”
Nucl. Fusion
40
,
1113
1123
(
2000
).
49.
C.
Bourdelle
,
Y.
Camenen
,
J.
Citrin
,
M.
Marin
,
F.
Casson
,
F.
Koechl
, and
M.
Maslov
, “
Fast H isotope and impurity mixing in ion-temperature-gradient turbulence
,”
Nucl. Fusion
58
,
076028
(
2018
).
You do not currently have access to this content.