Real-time control of a Tokamak plasma using neural networks

Christopher M. Bishop, P. S. Haynes, M. E. U. Smith, T. N. Todd, D. L. Trotman, C. G. Windsor

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

This paper presents results from the first use of neural networks for the real-time feedback control of high temperature plasmas in a Tokamak fusion experiment. The Tokamak is currently the principal experimental device for research into the magnetic confinement approach to controlled fusion. In the Tokamak, hydrogen plasmas, at temperatures of up to 100 Million K, are confined by strong magnetic fields. Accurate control of the position and shape of the plasma boundary requires real-time feedback control of the magnetic field structure on a time-scale of a few tens of microseconds. Software simulations have demonstrated that a neural network approach can give significantly better performance than the linear technique currently used on most Tokamak experiments. The practical application of the neural network approach requires high-speed hardware, for which a fully parallel implementation of the multi-layer perceptron, using a hybrid of digital and analogue technology, has been developed.
Original languageEnglish
Title of host publicationAdvances in Neural Information Processing Systems 7
EditorsG. Tesauro, D. S. Touretzky, T. D. Leen
Place of PublicationDever, US
PublisherMIT
Pages1007-1014
Number of pages8
Volume7
ISBN (Print)0262201046
Publication statusPublished - 28 Dec 1994
EventAdvances in Neural Information Processing Systems 1994 - Singapore, Singapore
Duration: 16 Nov 199418 Nov 1994

Other

OtherAdvances in Neural Information Processing Systems 1994
CountrySingapore
CitySingapore
Period16/11/9418/11/94

Fingerprint

feedback control
controlled fusion
self organizing systems
hydrogen plasma
high temperature plasmas
magnetic fields
hardware
fusion
high speed
analogs
computer programs
simulation
temperature

Bibliographical note

Copyright of the Massachusetts Institute of Technology Press (MIT Press)

Keywords

  • neural networks
  • temperature plasmas
  • Tokamak
  • fusion
  • magnetic
  • magnetic fields

Cite this

Bishop, C. M., Haynes, P. S., Smith, M. E. U., Todd, T. N., Trotman, D. L., & Windsor, C. G. (1994). Real-time control of a Tokamak plasma using neural networks. In G. Tesauro, D. S. Touretzky, & T. D. Leen (Eds.), Advances in Neural Information Processing Systems 7 (Vol. 7, pp. 1007-1014). Dever, US: MIT.
Bishop, Christopher M. ; Haynes, P. S. ; Smith, M. E. U. ; Todd, T. N. ; Trotman, D. L. ; Windsor, C. G. / Real-time control of a Tokamak plasma using neural networks. Advances in Neural Information Processing Systems 7. editor / G. Tesauro ; D. S. Touretzky ; T. D. Leen. Vol. 7 Dever, US : MIT, 1994. pp. 1007-1014
@inbook{63cbd4f7e8f04dceb7a15c447f6833cb,
title = "Real-time control of a Tokamak plasma using neural networks",
abstract = "This paper presents results from the first use of neural networks for the real-time feedback control of high temperature plasmas in a Tokamak fusion experiment. The Tokamak is currently the principal experimental device for research into the magnetic confinement approach to controlled fusion. In the Tokamak, hydrogen plasmas, at temperatures of up to 100 Million K, are confined by strong magnetic fields. Accurate control of the position and shape of the plasma boundary requires real-time feedback control of the magnetic field structure on a time-scale of a few tens of microseconds. Software simulations have demonstrated that a neural network approach can give significantly better performance than the linear technique currently used on most Tokamak experiments. The practical application of the neural network approach requires high-speed hardware, for which a fully parallel implementation of the multi-layer perceptron, using a hybrid of digital and analogue technology, has been developed.",
keywords = "neural networks, temperature plasmas, Tokamak, fusion, magnetic, magnetic fields",
author = "Bishop, {Christopher M.} and Haynes, {P. S.} and Smith, {M. E. U.} and Todd, {T. N.} and Trotman, {D. L.} and Windsor, {C. G.}",
note = "Copyright of the Massachusetts Institute of Technology Press (MIT Press)",
year = "1994",
month = "12",
day = "28",
language = "English",
isbn = "0262201046",
volume = "7",
pages = "1007--1014",
editor = "G. Tesauro and Touretzky, {D. S.} and Leen, {T. D.}",
booktitle = "Advances in Neural Information Processing Systems 7",
publisher = "MIT",

}

Bishop, CM, Haynes, PS, Smith, MEU, Todd, TN, Trotman, DL & Windsor, CG 1994, Real-time control of a Tokamak plasma using neural networks. in G Tesauro, DS Touretzky & TD Leen (eds), Advances in Neural Information Processing Systems 7. vol. 7, MIT, Dever, US, pp. 1007-1014, Advances in Neural Information Processing Systems 1994, Singapore, Singapore, 16/11/94.

Real-time control of a Tokamak plasma using neural networks. / Bishop, Christopher M.; Haynes, P. S.; Smith, M. E. U.; Todd, T. N.; Trotman, D. L.; Windsor, C. G.

Advances in Neural Information Processing Systems 7. ed. / G. Tesauro; D. S. Touretzky; T. D. Leen. Vol. 7 Dever, US : MIT, 1994. p. 1007-1014.

Research output: Chapter in Book/Report/Conference proceedingChapter

TY - CHAP

T1 - Real-time control of a Tokamak plasma using neural networks

AU - Bishop, Christopher M.

AU - Haynes, P. S.

AU - Smith, M. E. U.

AU - Todd, T. N.

AU - Trotman, D. L.

AU - Windsor, C. G.

N1 - Copyright of the Massachusetts Institute of Technology Press (MIT Press)

PY - 1994/12/28

Y1 - 1994/12/28

N2 - This paper presents results from the first use of neural networks for the real-time feedback control of high temperature plasmas in a Tokamak fusion experiment. The Tokamak is currently the principal experimental device for research into the magnetic confinement approach to controlled fusion. In the Tokamak, hydrogen plasmas, at temperatures of up to 100 Million K, are confined by strong magnetic fields. Accurate control of the position and shape of the plasma boundary requires real-time feedback control of the magnetic field structure on a time-scale of a few tens of microseconds. Software simulations have demonstrated that a neural network approach can give significantly better performance than the linear technique currently used on most Tokamak experiments. The practical application of the neural network approach requires high-speed hardware, for which a fully parallel implementation of the multi-layer perceptron, using a hybrid of digital and analogue technology, has been developed.

AB - This paper presents results from the first use of neural networks for the real-time feedback control of high temperature plasmas in a Tokamak fusion experiment. The Tokamak is currently the principal experimental device for research into the magnetic confinement approach to controlled fusion. In the Tokamak, hydrogen plasmas, at temperatures of up to 100 Million K, are confined by strong magnetic fields. Accurate control of the position and shape of the plasma boundary requires real-time feedback control of the magnetic field structure on a time-scale of a few tens of microseconds. Software simulations have demonstrated that a neural network approach can give significantly better performance than the linear technique currently used on most Tokamak experiments. The practical application of the neural network approach requires high-speed hardware, for which a fully parallel implementation of the multi-layer perceptron, using a hybrid of digital and analogue technology, has been developed.

KW - neural networks

KW - temperature plasmas

KW - Tokamak

KW - fusion

KW - magnetic

KW - magnetic fields

M3 - Chapter

SN - 0262201046

VL - 7

SP - 1007

EP - 1014

BT - Advances in Neural Information Processing Systems 7

A2 - Tesauro, G.

A2 - Touretzky, D. S.

A2 - Leen, T. D.

PB - MIT

CY - Dever, US

ER -

Bishop CM, Haynes PS, Smith MEU, Todd TN, Trotman DL, Windsor CG. Real-time control of a Tokamak plasma using neural networks. In Tesauro G, Touretzky DS, Leen TD, editors, Advances in Neural Information Processing Systems 7. Vol. 7. Dever, US: MIT. 1994. p. 1007-1014