Plasma and Fusion Research

Volume 17, 2402042 (2022)

Regular Articles

Data-Driven Control for Radiative Collapse Avoidance in Large Helical Device
Tatsuya YOKOYAMA1,2), Hiroshi YAMADA1), Suguru MASUZAKI3,4,5), Byron J. PETERSON3,4), Ryuichi SAKAMOTO1,3), Motoshi GOTO3,4), Tetsutaro OISHI3,4), Gakushi KAWAMURA3,4), Masahiro KOBAYASHI3,4), Toru I TSUJIMURA3), Yoshinori MIZUNO3), Junichi MIYAZAWA3,4), Kiyofumi MUKAI3,4), Naoki TAMURA3,4), Gen MOTOJIMA3,4) and Katsumi IDA3,4)
1)
Graduate School of Frontier Sciences, The University of Tokyo, Chiba 277-8561, Japan
2)
Research Fellow of Japan Society for the Promotion of Science, Tokyo 102-0083, Japan
3)
National Institute for Fusion Science, National Institutes of Natural Sciences, Gifu 509-5292, Japan
4)
The Graduate University for Advanced Studies, SOKENDAI, Gifu 509-5292, Japan
5)
Advanced Fusion Research Center, Research Institute for Applied Mechanics, Kyushu University, Fukuoka 816-8580, Japan
(Received 13 December 2021 / Accepted 6 March 2022 / Published 30 March 2022)

Abstract

A radiative collapse predictor has been developed using a machine-learning model with high-density plasma experiments in the Large Helical Device (LHD). The model is based on the collapse likelihood, which is quantified by the parameters selected by the sparse modeling, including ne, CIV, OV, and Te,edge. The control system implementing this model has been constructed with a single-board computer to apply this predictor model to the LHD experiment. The controller calculates the collapse likelihood and regulates gas-puff fueling and boosts electron cyclotron resonance heating in real-time. In density ramp-up experiments with hydrogen plasma, high-density plasma has been maintained by the control system while avoiding radiative collapse. This result has shown that the predictor based on the collapse likelihood has the capability to predict a radiative collapse in real-time.

Keywords

Large Helical Device (LHD), radiative collapse, density limit, stellarator-heliotron plasmas, sparse modeling, data-driven science, plasma control, collapse avoidance

DOI: 10.1585/pfr.17.2402042

Full Text

PDF format (2.3Mbytes) PDF Download

References

  • [1] S. Sudo et al., Nucl. Fusion 30(1), 11 (1990).
  • [2] A. Murari et al., Nucl. Fusion 60(5), 056003 (2020).
  • [3] C. Rea et al., Fusion Sci. Technol. 76(8), 912 (2020).
  • [4] K.J. Montes et al., Nucl. Fusion 61(2), 026022 (2021).
  • [5] J. Kates-Harbeck, A. Svyatkovskiy and W. Tang, Nature 568(7753), 526 (2019).
  • [6] C. Rea et al., Nucl. Fusion 59(9), 096016 (2019).
  • [7] W. Hu et al. Nucl. Fusion 61(6), 066034 (2021).
  • [8] T. Yokoyama et al., J. Fusion Energy 39(6), 500 (2020).
  • [9] T. Yokoyama et al., Plasma Fusion Res. 16, 2402010 (2021).
  • [10] A. Iiyoshi et al., Nucl. Fusion 39(9Y), 1245 (1999).
  • [11] C. Cortes and V. Vapnik, Machine Learning 20(3), 273 (1995).
  • [12] Y. Igarashi et al., J. Phys.: Conf. Series 699(1), 012001 (2016).
  • [13] C. Van Rijsbergen, Information Retrieval. (Butterworths, London, 1979).
  • [14] Y. Igarashi et al., J. Phys.: Conf. Series 1036, 012001 (2018).
  • [15] J. Miyazawa, K. Yasui and H. Yamada, Fusion Eng. Des. 83(2), 265 (2008). Proceedings of the 6th IAEA Technical Meeting on Control, Data Acquisition, and Remote Participation for Fusion Research.
  • [16] T.I. Tsujimura et al., Fusion Eng. Des. 153, 111480 (2020).
  • [17] T.I. Tsujimura et al., Nucl. Fusion 61(2), 026012 (2021).