Regular Articles

Full Text / References

Estimation of Plasma Vertical Position by Long-Short Term Memory Network with Time2Vec in a Small Tokamak Device PHiX

Sejung JANG and Hiroaki TSUTSUI

Tokyo Institute of Technology, Tokyo 152-8550, Japan

(Received 22 February 2024 / Accepted 2 June 2024 / Published 16 July 2024)

Abstract

Time evolution of plasma vertical position is estimated by using long-short term memory networks (LSTM) with Time2Vec technique which incorporates temporal information into a neural network. Since many tokamak devices have elongated cross-section in achieving high performance whereas accurate vertical position feedback control is required in order to avoid vertical displacement events (VDEs). Our data-driven model, using experimental data obtained from a small tokamak device PHiX in Tokyo Institute of Technology, can estimate the plasma vertical displacement by incorporating operational scenario coils current data. The model achieved high performance by combining Time2Vec with LSTM. We can also interpret the weights extracted from a trained, data-driven model by comparing the model’s predictions.

Copyright (c) 2024 The Japan Society of Plasma Science and Nuclear Fusion Research

Keywords

tokamak, vertical instability, LSTM, machine learning, Time2Vec

DOI: 10.1585/pfr.19.1403023

Full Text

PDF format (1.0Mbytes)

References

• [1] F. Troyon, R. Gruber, H. Saurenmann, S. Semenzato and S. Succi, Plasma Phys. Control. Fusion 26(1)A, 209 (1984).
• [2] R.J. Goldston, Plasma Phys. Control. Fusion 26(1)A, 87 (1984).
• [3] Y. Nakamura, R. Yoshino, N. Pomphrey and S.C. Jardin, J. Nucl. Sci. Technol. 33(8), 609 (1996).
• [4] S. Weidman, Deep Learning from Scratch (O’Reilly Media, Sebastopol, CA, 2019), Chap. 2.
• [5] T. Yokoyama, H. Yamada, S. Masuzaki, B.J. Peterson, R. Sakamoto, M. Goto et al., Plasma Fusion Res. 17, 2402042 (2022).
• [6] S. Inoue, Y. Miyata, H. Urano and T. Suzuki, Nucl. Fusion 62(8), 086007 (2022).
• [7] D.E. Rumelhart, G.E. Hinton and R.J. Williams, Nature 323(6088), 533 (1986).
• [8] S. Hochreiter and J. Schmidhuber, Neural Comput. 9(8), 1735 (1997).
• [9] J. Kates-Harbeck, A. Svyatkovskiy and W. Tang, Nature 568, 526 (2019).
• [10] A. Pau, A. Fanni, S. Carcangiu, B. Cannas, G. Sias, A. Murari and F. Rimini, Nucl. Fusion 59, 106017 (2019).
• [11] A. Agarwal, A. Mishra, P. Sharma, S. Jain, S. Ranjan and R. Manchanda, Plasma Phys. Control. Fusion 63(11), 115004 (2021).
• [12] S.M. Kazemi, R. Goel, S. Eghbali, J. Ramanan, J. Sahota, S. Thakur et al., arXiv preprint arXiv:1907.05321 (2019).
• [13] D.P. Kingma and J. Ba, arXiv preprint arXiv:1412.6980 (2014).
• [14] A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan et al., Advances in Neural Information Processing Systems 32, 7994 (2019).
• [15] S. Naito, M. Murayama, S. Hatakeyama, D. Kuwahara, Y. Suzuki, H. Tsutsui and S. Tsuji-Iio, Nucl. Fusion 61(11), 116035 (2021).
• [16] K. Munechika, H. Tsutsui and S. Tsuji-Iio, Plasma Fusion Res. 16, 2402033 (2021).
• [17] A. Reiman, Phys. Rev. 99, 135007 (2007).
• [18] K. Yasuda, T. Fujita, A. Okamoto, H. Arimoto, S. Kimata, K. Kado and K. Tsunoda, Phys. Plasmas 28, 082108 (2021).
• [19] S. Naito, Y. Suzuki and H. Tsutsui, Plasma Fusion Res. 17, 2403069 (2022).
• [20] K. Kurihara, Y. Kawamata, M. Sueoka, H. Hosoyama, I. Yonekawa, T. Suzuki, T. Oikawa, S. Ide and JT-60 Team, Fusion Eng. Des. 74(1-4), 527 (2005).