Disruption prediction on EAST tokamak using a deep learning algorithm
Authors:
- B H GuoBGB H Guo
- D L ChenDCD L Chen
- B ShenBSB Shen
- C ReaCRC Rea
- R S GranetzRGR S Granetz
- L ZengLZL Zeng
- W H HuWHW H Hu
- J P QianJQJ P Qian
- Y W SunYSY W Sun
- B J XiaoBXB J Xiao
- 24 September 2021
- journal article
- research article
- Published by IOP Publishing in Plasma Physics and Controlled Fusion
- Vol. 63 (11) , 115007
Abstract
In this study, a long short-term memory (LSTM) model is trained on a large disruption warning database to predict the disruption on EAST tokomak. To compare the performance of the proposed model with the previously reported full convolutional neural network (CNN) (Guo et al 2020 Plasma Phys. Control. Fusion 63 025008), the same data set and diagnostic signals are used. Based on the test set, the area under the receiver operating characteristic curve, i.e. the AUC value of the LSTM model is obtained as 0.87, and the true positive rate (TPR) is sim87.5%, while the false positive rate (FPR) is sim15.1%. Since the LSTM model is more sensitive to radiation fluctuations than CNN, the prediction performance of LSTM model is inferior to that of CNN model (for CNN, AUC sim 0.92, TPR sim 87.5%, FPR sim 6.1%). However, the advance warning time of LSTM model is 14 ms earlier than that of CNN. To reduce the FPR and improve the performance of the model, more fast bolometer channels are added as the input signals of the LSTM model, including the radiation from the upper and lower edges and the plasma core. Consequently, for the same test set, the AUC value increases to 0.89, and the FPR decreases to sim9.4%, but the TPR also decreases to sim83.9%. In addition, the sensitivity of the model to radiation fluctuations caused by impurity behavior decreases significantly, and the warning time becomes 8.7 ms earlier as compared to that of the original model. Overall, it is proved that deep learning algorithms exhibit immense application potential in the disruption prediction of long-pulse fusion devices.Keywords
Funding Information
- Youth Innovation Promotion Association of Chinese Academy of Sciences (2017480)
- National MCF Energy R&D Program of China (2018YFE0302100)
- National Natural Science Foundation of China (11775262)
This publication has 23 references indexed in Scilit:
- Hybrid neural network for density limit disruption prediction and avoidance on J-TEXT tokamakNuclear Fusion, 2018
- ImageNet classification with deep convolutional neural networksCommunications of the ACM, 2017
- Establishment and Assessment of Plasma Disruption and Warning Databases from EASTPlasma Science and Technology, 2016
- Requirements for Triggering the ITER Disruption Mitigation SystemFusion Science and Technology, 2016
- Study of the radiated energy loss during massive gas injection mitigated disruptions on EASTJournal of Nuclear Materials, 2015
- Implementation of the Disruption Predictor APODIS in JET’s Real-Time Network Using the MARTe FrameworkIEEE Transactions on Nuclear Science, 2014
- Development of an efficient real-time disruption predictor from scratch on JET and implications for ITERNuclear Fusion, 2013
- Neural-net predictor for beta limit disruptions in JT-60UNuclear Fusion, 2005
- Chapter 3: MHD stability, operational limits and disruptionsNuclear Fusion, 1999
- Tokamak disruption alarm based on a neural network model of the high- beta limitNuclear Fusion, 1997