Hybrid deep learning architecture for general disruption prediction across tokamaks

Zhu, J. X.; Rea, C.; Montes, K.; Granetz, R. S.; Sweeney, R.; Tinguely, R. A.

doi:10.7910/DVN/XIOHW1

Title: Hybrid deep learning architecture for general disruption prediction across tokamaks

Dataset
Other Related Research

Abstract

In this paper, we present a new deep learning disruption prediction algorithm based on important findings from explorative data analysis which effectively allows knowledge transfer from existing devices to new ones, thereby predicting disruptions using very limited disruptive data from the new devices. The explorative data analysis conducted via unsupervised clustering techniques confirms that time-sequence data are much better separators of disruptive and non-disruptive behavior than the instantaneous plasma state data with further advantageous implications for a sequence-based predictor. Based on such important findings, we have designed a new algorithm for multi-machine disruption prediction that achieves high predictive accuracy on the C-Mod (AUC=0.801), DIII-D (AUC=0.947) and EAST (AUC=0.973). tokamaks with limited hyperparameter tuning. Through numerical experiments, we show that boosted accuracy (AUC=0.959) is achieved on EAST predictions by including in the training only 20 disruptive discharges, thousands of non-disruptive discharges from EAST, and combining this with more than a thousand discharges from DIII-D and C-Mod. The improvement of predictive ability obtained by combining disruptive data from other devices is found to be true for all permutations of the three devices. Furthermore, by comparing the predictive performance of each individual numerical experiment, we find that non-disruptive data are machine-specific while disruptivemore »« less

Authors:

Zhu, J. X.; Rea, C.; Montes, K.; Granetz, R. S.; Sweeney, R.; Tinguely, R. A.

OSTI

Publication Date:: Fri Jun 25 00:00:00 EDT 2021

DOE Contract Number:: FC02-04ER54698; SC0014264

Research Org.:: Massachusetts Inst. of Technology (MIT), Cambridge, MA (United States). Plasma Science and Fusion Center

Sponsoring Org.:: USDOE Office of Science (SC), Fusion Energy Sciences (FES)

Subject:: 70 PLASMA PHYSICS AND FUSION TECHNOLOGY

OSTI Identifier:: 1886671

DOI:: https://doi.org/10.7910/DVN/XIOHW1

Citation Formats


                    Zhu, J. X., Rea, C., Montes, K., Granetz, R. S., Sweeney, R., and Tinguely, R. A. Hybrid deep learning architecture for general disruption prediction across tokamaks.  United States: N. p., 2021. 
        Web.  doi:10.7910/DVN/XIOHW1.

Copy to clipboard


                    Zhu, J. X., Rea, C., Montes, K., Granetz, R. S., Sweeney, R., & Tinguely, R. A. Hybrid deep learning architecture for general disruption prediction across tokamaks.  United States.  doi:https://doi.org/10.7910/DVN/XIOHW1

Copy to clipboard


                    Zhu, J. X., Rea, C., Montes, K., Granetz, R. S., Sweeney, R., and Tinguely, R. A. 2021.  
"Hybrid deep learning architecture for general disruption prediction across tokamaks".  United States.  doi:https://doi.org/10.7910/DVN/XIOHW1.  https://www.osti.gov/servlets/purl/1886671. Pub date:Fri Jun 25 00:00:00 EDT 2021

Copy to clipboard


                    
@article{osti_1886671,

  title        = {Hybrid deep learning architecture for general disruption prediction across tokamaks},

  author       = {Zhu, J. X. and Rea, C. and Montes, K. and Granetz, R. S. and Sweeney, R. and Tinguely, R. A.},

  abstractNote = {In this paper, we present a new deep learning disruption prediction algorithm based on important findings from explorative data analysis which effectively allows knowledge transfer from existing devices to new ones, thereby predicting disruptions using very limited disruptive data from the new devices. The explorative data analysis conducted via unsupervised clustering techniques confirms that time-sequence data are much better separators of disruptive and non-disruptive behavior than the instantaneous plasma state data with further advantageous implications for a sequence-based predictor. Based on such important findings, we have designed a new algorithm for multi-machine disruption prediction that achieves high predictive accuracy on the C-Mod (AUC=0.801), DIII-D (AUC=0.947) and EAST (AUC=0.973). tokamaks with limited hyperparameter tuning. Through numerical experiments, we show that boosted accuracy (AUC=0.959) is achieved on EAST predictions by including in the training only 20 disruptive discharges, thousands of non-disruptive discharges from EAST, and combining this with more than a thousand discharges from DIII-D and C-Mod. The improvement of predictive ability obtained by combining disruptive data from other devices is found to be true for all permutations of the three devices. Furthermore, by comparing the predictive performance of each individual numerical experiment, we find that non-disruptive data are machine-specific while disruptive data from multiple devices contain device-independent knowledge that can be used to inform predictions for disruptions occurring on a new device.},

  doi          = {10.7910/DVN/XIOHW1},

  journal      = {},

  number       = ,

  volume       = ,

  place        = {United States},

  year         = {Fri Jun 25 00:00:00 EDT 2021},

  month        = {Fri Jun 25 00:00:00 EDT 2021}

}

Copy to clipboard

Dataset:

View Dataset

DOI: https://doi.org/10.7910/DVN/XIOHW1

Save / Share:

Export Metadata

Save to My Library

Similar records in DOE Data Explorer and OSTI.GOV collections:

Hybrid deep learning architecture for general disruption prediction across tokamaks

Journal Article Zhu, Jinxiang ; Rea, Cristina ; Montes, Kevin Joseph ; ... - Nuclear Fusion

In this paper, we present a new deep learning disruption prediction algorithm based on important findings from explorative data analysis which effectively allows knowledge transfer from existing devices to new ones, thereby predicting disruptions using very limited disruptive data from the new devices. Here, the explorative data analysis conducted via unsupervised clustering techniques confirms that time-sequence data are much better separators of disruptive and non-disruptive behavior than the instantaneous plasma state data with further advantageous implications for a sequence-based predictor. Based on such important findings, we have designed a new algorithm for multi-machine disruption prediction that achieves high predictive accuracymore »« less
Progress Towards Interpretable Machine Learning-based Disruption Predictors Across Tokamaks

Dataset Rea, C. ; Montes, K. J. ; Pau, A. ; ...

In this paper we lay the groundwork for a robust cross-device comparison of data-driven disruption prediction algorithms on DIII-D and JET tokamaks. In order to consistently carry on a comparative analysis, we define physics-based indicators of disruption precursors based on temperature, density, and radiation profiles that are currently missing for DIII-D data. These profile-based indicators are shown to well-describe impurity accumulation events in both DIII-D and JET discharges that eventually disrupt. Thanks to the univariate analysis on the features used in such data-driven applications on both tokamaks, we are able to statistically highlight differences in the dominant disruption precursors: JETmore »« less
Density Limits as Disruption Forecasters for Spherical Tokamaks

Dataset Berkery, John
Deep convolutional neural networks for multi-scale time-series classification and application to disruption prediction in fusion devices

Dataset Churchill, R M

The multi-scale, mutli-physics nature of fusion plasmas makes predicting plasma events challenging. Recent advances in deep convolutional neural network architectures (CNN) utilizing dilated convolutions enable accurate predictions on sequences which have long-range, multi-scale characteristics, such as the time-series generated by diagnostic instruments observing fusion plasmas. Here we apply this neural network architecture to the popular problem of disruption prediction in fusion tokamaks, utilizing raw data from a single diagnostic, the Electron Cyclotron Emission imaging (ECEi) diagnostic from the DIII-D tokamak. ECEi measures a fundamental plasma quantity (electron temperature) with high temporal resolution over the entire plasma discharge, making it sensitive to a number of potential pre-disruptions markers with different temporal and spatial scales. Promising, initial disruption prediction results are obtained training a deep CNN with large receptive field ({more »« less
Electromagnetic Particle Injector for Fast Time Response Disruption Mitigation in Tokamaks

Dataset Raman, R ; Lay, W -S ; Jarboe, T R ; ...

A novel, rapid time-response, disruption mitigation system referred to as the Electromagnetic Particle Injector (EPI) is described. This method can accurately deliver the radiative payload to the plasma center on a <10 ms time scale, much faster, and deeper, than what can be achieved using conventional methods. The EPI system accelerates a sabot electromagnetically. The sabot is a metallic capsule that can be accelerated to desired velocities by an electromagnetic impeller. At the end of its acceleration, within 2 ms, the sabot will release a radiative payload, which is composed of low-z granules, or a shell pellet containing smaller pellets.more »« less

Similar Records