Recovery of electron density signals beyond the operating range of the measuring instrument

封面

如何引用文章

全文:

开放存取 开放存取
受限制的访问 ##reader.subscriptionAccessGranted##
受限制的访问 订阅或者付费存取

详细

Machine learning models have been widely incorparated into control systems aimed at improving the operational efficiency of tokamaks. The training machine learning models requires substantial datasets. However, data collection is limited because experimental campaigns on tokamaks are prolonged in time. Furthermore, the amount of suitable training data may decrease due to the present of faulty diagnostic signals. Additionally, the frequency of faulty signal occurrences increases while initial operation of a new tokamak or specialized equipment. This work examines the possibility of recovering faulty signals using machine learning techniques. Particularly, we focus on recovering signals obtained beyond the operating range of measuring instruments. Thus, recovering such kind of signals should increase the volume of available training data, consequently enhancing the efficacy of machine learning-based model training.

全文:

受限制的访问

作者简介

Nikolai Leshov

Moscow Technical University of Communications and Informatics (MTUCI); State Research Centre of the Russian Federation Troitsk Institute for Innovation and Fusion Research

编辑信件的主要联系方式.
Email: nikolya.leshov@gmail.com
ORCID iD: 0000-0002-7844-1768

postgraduate student, Department of Mathematical Cybernetics and Information Technologies

俄罗斯联邦, Moscow; Troitsk, Moscow

Anastasia Shcherbak

State Research Centre of the Russian Federation Troitsk Institute for Innovation and Fusion Research

Email: shcherbak@triniti.ru
ORCID iD: 0000-0002-0942-9837

leading engineer, Laboratory of Tokamak Plasma Diagnostics and Plasma Physics, Department of Tokamak and Current-Carrying Plasma Physics

俄罗斯联邦, Troitsk, Moscow

Mikhail Gorodnichev

Moscow Technical University of Communications and Informatics (MTUCI)

Email: m.g.gorodnichev@mtuci.ru
ORCID iD: 0000-0003-1739-9831

Cand. Sci. (Eng.), Associate Professor, Dean, Faculty “Information Technology”

俄罗斯联邦, Moscow

参考

  1. Wesson J. Tokamaks. 4th ed. Oxford University Press, 2011. 828 p. (International Series of Monographs on Physics)
  2. O’Shea F. H. et al. Coincidence anomaly detection for unsupervised locating of edge localized modes in the DIII-D tokamak dataset. Machine Learning: Science and Technology. 2024. Vol. 5. No. 3. 035050.
  3. Lu J. et al. Fast equilibrium reconstruction by deep learning on EAST tokamak. AIP Advances. 2023. Vol. 13. No. 7.
  4. Degrave J. et al. Magnetic control of tokamak plasmas through deep reinforcement learning. Nature. 2022. Vol. 602. No. 7897. Pp. 414–419.
  5. Zheng W. et al. Hybrid neural network for density limit disruption prediction and avoidance on J-TEXT tokamak. Nuclear Fusion. 2018. Vol. 58. No. 5. 056016.
  6. Zhu J. X. et al. Integrated deep learning framework for unstable event identification and disruption prediction of tokamak plasmas. Nuclear Fusion. 2023. Vol. 63. No. 4. 046009.
  7. Zhu J. X. et al. Hybrid deep-learning architecture for general disruption prediction across multiple tokamaks. Nuclear Fusion. 2020. Vol. 61. No. 2. 026007.
  8. Yang Z. et al. Implementing deep learning-based disruption prediction in a drifting data environment of new tokamak: HL-3. Nuclear Fusion. 2024.
  9. Abbate J. et al. Data-driven profile prediction for DIII-D. Nuclear Fusion. 2021. Vol. 61. No. 4. 046027.
  10. Felici F. et al. Real-time-capable prediction of temperature and density profiles in a tokamak using RAPTOR and a first-principle-based transport model. Nuclear Fusion. 2018. Vol. 58. No. 9. 096006.
  11. Chayapathy D. et al. Time series viewmakers for robust disruption prediction. In: Machine learning and the physical sciences workshop. NeurIPS, 2024.
  12. Hochreiter S. et al. Long short-term memory. Neural Computation. 1997. Vol. 9. No. 8. Pp. 1735–1780.
  13. Guo B.H. et al. Disruption prediction on EAST tokamak using a deep learning algorithm. Plasma Physics and Controlled Fusion. 2021. Vol. 63. No. 11. 115007.
  14. Seo J. et al. Feedforward beta control in the KSTAR tokamak by deep reinforcement learning. Nuclear Fusion. 2021. Vol. 61. No. 10. 106010.
  15. Matos F. et al. Classification of tokamak plasma confinement states with convolutional recurrent neural networks. Nuclear Fusion. 2020. Vol. 60. No. 3. 036022.
  16. Akiba T. et al. Optuna: A next-generation hyperparameter optimization framework. In: Proceedings of the 25th ACM SIGKDD international conference on knowledge discovery & data mining. 2019. Pp. 2623–2631.
  17. Kingma D. P. et al. Adam: A method for stochastic optimization. Published as a conference paper at the 3rd International Conference for Learning Representations. San Diego, 2015.
  18. Paszke A. et al. Pytorch: An imperative style, high-performance deep learning library. In: Advances in neural information processing systems. 2019. Vol. 32.

补充文件

附件文件
动作
1. JATS XML
下载
2. Fig. 1. Correct signal

下载 (144KB)
索引源数据
3. Fig. 2. Signal beyond operating range

下载 (139KB)
索引源数据
4. Fig. 3. Transformed signal beyond operating range

下载 (140KB)
索引源数据
5. Fig. 4. Distribution by duration of the “gap”

下载 (189KB)
索引源数据
6. Fig. 5. Synthetic signal

下载 (136KB)
索引源数据
7. Fig. 6. Distribution of signals beyond operating range in space

下载 (243KB)
索引源数据
8. Fig. 7. Distribution of correct signals in space

下载 (394KB)
索引源数据
9. Fig. 8. Model architecture

下载 (111KB)
索引源数据
10. Fig. 9. Learning curve

下载 (116KB)
索引源数据
11. Fig. 10. Recovered synthetic signal

下载 (152KB)
索引源数据
12. Fig. 11. Recovered signal and original signal

下载 (597KB)
索引源数据

版权所有 © Yur-VAK, 2025

许可 URL: https://www.urvak.ru/contacts/