Detection of Disruptions in the High-β Spherical Torus NSTX

Gerhardt, S P; Bell, R E; LeBlanc, B P; Menard, J E; Mueller, D; Roquemore, A L; Sabbagh, S A

doi:10.2172/1062404

Cited by 6 publications

(10 citation statements)

References 135 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Machine-learning approaches (discussed below) should be applicable to data analysis for optimization of these individual physics-based tests [33] and the control responses. Although each such test is specialized to certain classes of disruptions, multiple threshold tests can be combined with suitable weighting to make a more general warning system with a high success rate [34].…”

Section: Real-time Predictionmentioning

confidence: 99%

Progress in disruption prevention for ITER

Strait¹,

Barr²,

Baruzzo³

et al. 2019

Nucl. Fusion

View full text Add to dashboard Cite

Key plasma physics and real-time control elements needed for robustly stable operation of high fusion power discharges in ITER have been demonstrated in US fusion research. Optimization of the current density profile has enabled passively stable operation without n " 1 tearing modes in discharges simulating ITER's baseline scenario with zero external torque. Stable rampdown of the discharge has been achieved with ITER-like scaled current ramp rates, while maintaining an X-point configuration. Significant advances have been made toward real-time prediction of disruptions: machine learning techniques for prediction of disruptions have achieved 90% accuracy in offline analysis, and direct probing of ideal and resistive plasma stability using 3D magnetic perturbations has shown a rising plasma response before the onset of a tearing mode. Active stability control contributes to prevention of disruptions, including direct stabilization of resistive-wall kink modes in high-β discharges, forced rotation of magnetic islands to prevent wall locking, and localized heating/current drive to shrink the islands. These elements are being integrated into stable operating scenarios and a new event-handling system for off-normal events in order to develop the physics basis and techniques for robust control in ITER.

show abstract

Section: Real-time Predictionmentioning

confidence: 99%

Progress in disruption prevention for ITER

Strait¹,

Barr²,

Baruzzo³

et al. 2019

Nucl. Fusion

View full text Add to dashboard Cite

show abstract

“…There are many approaches to create such algorithms. Perhaps the simplest is a scheme in which thresholds (of physical quantities in parameter space) between classes ND and D are determined by minimizing overlap in histograms of data from each class, as discussed in [5,11,23,28] among other works. Machine learning methods have also been implemented, trained and tested on these data sets (see the references listed in section 1).…”

Section: Interpretation Of Random Forest Predictionsmentioning

confidence: 99%

“…differences between programmed and measured values of plasma current or vertical position-are chosen based on their relevance to disruption physics as well as their real-time measurement capability. Second, databases of these parameters are assembled for many machines and many times throughout plasma discharges, some terminated by disruptions; for examples, see [7,10,11,13,17,28].…”

Section: Implication Of Binary Classificationmentioning

confidence: 99%

“…However, for both avoidance and mitigation, an impending disruption must first be predicted with enough time to actuate an appropriate response. There are many past and ongoing efforts to develop disruption prediction algorithms, including work on tokamaks ADITYA [3,4], Alcator C-Mod [5,6], ASDEX-U [7][8][9], DIII-D [5,6,[10][11][12], EAST [6], JET [8,[13][14][15][16][17][18][19][20][21][22][23], JT60-U [24,25], J-TEXT [26,27], and NSTX [28], among others. At present, most predictors estimate the plasma state at each moment in time, usually identified as either non-disruptive or disruptive.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

An application of survival analysis to disruption prediction via Random Forests

Tinguely

Montes

Rea

et al. 2019

Plasma Phys. Control. Fusion

View full text Add to dashboard Cite

One of the most pressing challenges facing the fusion community is adequately mitigating or, even better, avoiding disruptions of tokamak plasmas. However, before this can be done, disruptions must first be predicted with sufficient warning time to actuate a response. The established field of survival analysis provides a convenient statistical framework for time-to-event (i.e. time-to-disruption) studies. This paper demonstrates the integration of an existing disruption prediction machine learning algorithm with the Kaplan-Meier estimator of survival probability. Specifically discussed are the implied warning times from binary classification of disruption databases and the interpretation of output signals from Random Forest algorithms trained and tested on these databases. This survival analysis approach is applied to both smooth and noisy test data to highlight important features of the survival and hazard functions. In addition, this method is applied to three Alcator C-Mod plasma discharges and compared to a threshold-based scheme for triggering alarms. In one case, both techniques successfully predict the disruption; although, in another, neither warns of the impending disruption with enough time to mitigate. For the final discharge, the survival analysis approach could avoid the false alarm triggered by the threshold method. Limitations of this analysis and opportunities for future work are also presented.

show abstract

“…In many toroidal plasma devices such as tokamaks and spherical toruses (STs) (Gerhardt et al, 2013), disruptions are observed as sudden and dangerous events that induce rapid release of particles and energy to the device wall (Schuller, 1995). A typical disruption brings the plasma experiment (the "shot") to an abrupt end and, because of the associated large rapid energy release, it can also seriously damage the device-especially in larger systems such as ITER (Strait et al, 2019).…”

Section: Introductionmentioning

confidence: 99%

Fully Convolutional Spatio-Temporal Models for Representation Learning in Plasma Science

Dong

Felker

Svyatkovskiy

et al. 2021

J Mach Learn Model Comput

View full text Add to dashboard Cite

We have trained a fully convolutional spatio-temporal model for fast and accurate representation learning in the challenging exemplar application area of fusion energy plasma science. The onset of major disruptions is a critically important fusion energy science issue that must be resolved for advanced tokamak plasmas such as the $25B burning plasma international thermonuclear experimental reactor (ITER) experiment. While a variety of statistical methods have been used to address the problem of tokamak disruption prediction and control, recent approaches based on deep learning have proven particularly compelling. In the present paper, we introduce further improvements to the fusion recurrent neural network (FRNN) software suite, which delivered cross-machine disruption predictions with unprecedented accuracy using a large database of experimental signals from two major tokamaks. Up to now, FRNN was based on the long short-term memory (LSTM) variant of recurrent neural networks to leverage the temporal information in the data. Here, we implement and apply the "temporal convolutional neural network (TCN)" architecture to the time-dependent input signals. This allows highly optimized convolution operations to carry the majority of the computational load of training, thus enabling a reduction in training time, and the effective use of highperformance computing resources for hyperparameter tuning. At the same time, the TCN-based architecture achieves better predictive performance when compared with the LSTM architecture for various tasks for a representative fusion database.

show abstract

Detection of Disruptions in the High-β Spherical Torus NSTX

Cited by 6 publications

References 135 publications

Progress in disruption prevention for ITER

Progress in disruption prevention for ITER

An application of survival analysis to disruption prediction via Random Forests

Fully Convolutional Spatio-Temporal Models for Representation Learning in Plasma Science

Contact Info

Product

Resources

About