Data-driven profile prediction for DIII-D

J. Abbate; R. Conlin; E. Kolemen

doi:10.1088/1741-4326/abe08d

Data-driven profile prediction for DIII-D

J. Abbate, R. Conlin, E. Kolemen

Research output: Contribution to journal › Article › peer-review

22 Scopus citations

Abstract

A new, fully data-driven algorithm has been developed that uses a neural network to predict plasma profiles on a scale of τ E into the future given an actuator trajectory and the plasma state history. The model was trained and tested on DIII-D data from the 2013-2018 experimental campaigns. The model runs in tens of milliseconds and is very simple to use. This makes it a potentially useful tool for operators and physicists when planning plasma scenarios. It is also fast enough to be used for real-time model-predictive control.

Original language	English (US)
Article number	046027
Journal	Nuclear Fusion
Volume	61
Issue number	4
DOIs	https://doi.org/10.1088/1741-4326/abe08d
State	Published - Apr 2021

All Science Journal Classification (ASJC) codes

Nuclear and High Energy Physics
Condensed Matter Physics

Keywords

DIII-D
control
neural networks
transport

Access to Document

10.1088/1741-4326/abe08d

Cite this

@article{88813989e1a543c29a4ed2b00da270be,

title = "Data-driven profile prediction for DIII-D",

abstract = "A new, fully data-driven algorithm has been developed that uses a neural network to predict plasma profiles on a scale of τ E into the future given an actuator trajectory and the plasma state history. The model was trained and tested on DIII-D data from the 2013-2018 experimental campaigns. The model runs in tens of milliseconds and is very simple to use. This makes it a potentially useful tool for operators and physicists when planning plasma scenarios. It is also fast enough to be used for real-time model-predictive control.",

keywords = "DIII-D, control, neural networks, transport",

author = "J. Abbate and R. Conlin and E. Kolemen",

note = "Publisher Copyright: {\textcopyright} 2021 IAEA, Vienna.",

year = "2021",

month = apr,

doi = "10.1088/1741-4326/abe08d",

language = "English (US)",

volume = "61",

journal = "Nuclear Fusion",

issn = "0029-5515",

publisher = "IOP Publishing Ltd.",

number = "4",

}

TY - JOUR

T1 - Data-driven profile prediction for DIII-D

AU - Abbate, J.

AU - Conlin, R.

AU - Kolemen, E.

PY - 2021/4

Y1 - 2021/4

N2 - A new, fully data-driven algorithm has been developed that uses a neural network to predict plasma profiles on a scale of τ E into the future given an actuator trajectory and the plasma state history. The model was trained and tested on DIII-D data from the 2013-2018 experimental campaigns. The model runs in tens of milliseconds and is very simple to use. This makes it a potentially useful tool for operators and physicists when planning plasma scenarios. It is also fast enough to be used for real-time model-predictive control.

AB - A new, fully data-driven algorithm has been developed that uses a neural network to predict plasma profiles on a scale of τ E into the future given an actuator trajectory and the plasma state history. The model was trained and tested on DIII-D data from the 2013-2018 experimental campaigns. The model runs in tens of milliseconds and is very simple to use. This makes it a potentially useful tool for operators and physicists when planning plasma scenarios. It is also fast enough to be used for real-time model-predictive control.

KW - DIII-D

KW - control

KW - neural networks

KW - transport

UR - http://www.scopus.com/inward/record.url?scp=85100621591&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85100621591&partnerID=8YFLogxK

U2 - 10.1088/1741-4326/abe08d

DO - 10.1088/1741-4326/abe08d

M3 - Article

AN - SCOPUS:85100621591

SN - 0029-5515

VL - 61

JO - Nuclear Fusion

JF - Nuclear Fusion

IS - 4

M1 - 046027

ER -

Data-driven profile prediction for DIII-D

Abstract

All Science Journal Classification (ASJC) codes

Keywords

Access to Document

Other files and links

Fingerprint

Cite this