A hybrid physics/data-driven logic to detect, classify, and predict anomalies and disruptions in tokamak plasmas

Rossi, R.; Gelfusa, M.; Craciunescu, T.; Wyss, I.; Vega, J.; ,

doi:10.1088/1741-4326/ad2723

Cited by 3 publications

(1 citation statement)

References 70 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Disruption prediction models, including this work, aim to be used for disruption avoidance [61][62][63] or mitigation [64]. This work focused on developing a reliable event alarm to predict the occurrence of disruptions with the Bayesian framework.…”

Section: Discussionmentioning

confidence: 99%

Enhancing disruption prediction through Bayesian neural network in KSTAR

Kim,

Lee,

Seo

et al. 2024

Plasma Phys. Control. Fusion

View full text Add to dashboard Cite

In this research, we develop a data-driven disruption predictor based on Bayesian deep probabilistic learning, capable of predicting disruptions and modeling uncertainty in KSTAR. Unlike conventional neural networks within a frequentist approach, Bayesian neural networks can quantify the uncertainty associated with their predictions, thereby enhancing the precision of disruption prediction by mitigating false alarm rates through uncertainty thresholding. Leveraging 0D plasma parameters from EFIT and diagnostic data, a Temporal Convolutional Network adept at handling multi-time scale data was utilized. The proposed framework demonstrates proficiency in predicting disruptions, substantiating its effectiveness through successful applications to KSTAR experimental data.

show abstract

Section: Discussionmentioning

confidence: 99%

Enhancing disruption prediction through Bayesian neural network in KSTAR

Kim,

Lee,

Seo

et al. 2024

Plasma Phys. Control. Fusion

View full text Add to dashboard Cite

show abstract

An advanced double-phase stacking ensemble technique with active learning classifier: Toward reliable disruption prediction in Aditya tokamak

Muruganandham,

Jayaraman,

Tahiliani

et al. 2024

Review of Scientific Instruments

View full text Add to dashboard Cite

Disruptions in tokamak nuclear reactors, where plasma confinement is suddenly lost, pose a serious threat to the reactor and its components. Classifying discharges as disruptive or non-disruptive is crucial for effective plasma operation and advanced prediction. Traditional disruption identification systems often struggle with noise, variability, and limited adaptability. To address these challenges, we propose an enhanced stacking generalization model called the “Double-Phase Stacking Technique” integrated with Pool-based Active Learning (DPST-PAL) for designing a robust classifier with minimal labor cost. This innovative approach improves classification accuracy and reliability using advanced data analysis techniques. We trained the DPST-PAL model on 162 diagnostic shots from the Aditya dataset, achieving a high accuracy of 98% and an F1-score of 0.99, surpassing conventional methods. Subsequently, the deep 1D convolutional predictor model is implemented and trained using the classified shots obtained from the DPST-PAL model to validate the reliability of the dataset, which is tested on 47 distinct shots. This model accurately predicts the disruptions 7–13 ms in advance with 93.6% accuracy and exhibited no premature alarms or misclassifications for our experimental shots.

show abstract

A self-organised partition of the high dimensional plasma parameter space for plasma disruption prediction

Aymerich,

Fanni,

Pisano

et al. 2024

Nucl. Fusion

View full text Add to dashboard Cite

This paper introduces a disruption predictor constructed through a fully unsupervised two-dimensional mapping of the high-dimensional JET operational space. The primary strength of this disruption predictor lies in its inherent self-organization capability. Diverging from both supervised disruption predictors and earlier approaches suggested by the same authors, which were based on unsupervised models such as Self-Organizing or Generative Topographic Maps, this predictor eliminates the need for labeling data of disruption terminated pulses during training. In prior methods, labels were indeed required post-mapping to inform the model about the presence or absence of disruption precursors at each time instant during the disrupted discharges. In contrast, our approach in this study involves no labeling of data from disruption-terminated experiments. The Self-Organizing Map, operating without any a priori information, adeptly identifies the regions characterizing the pre-disruptive phase. Moreover, SOM discovers non-trivial relationships and captures the complicated interplay of device diagnostics on the internal plasma states from the experimental data. The provided model is highly interpretable; it allows the visualization of high-dimensional data and facilitates easy interrogation of the model to understand the reasons behind its correlations. Hence, utilizing SOMs across various devices can prove invaluable in extracting rules and identifying common patterns, thereby facilitating extrapolation to ITER of the knowledge acquired from existing tokamaks.

show abstract

A hybrid physics/data-driven logic to detect, classify, and predict anomalies and disruptions in tokamak plasmas

Cited by 3 publications

References 70 publications

Enhancing disruption prediction through Bayesian neural network in KSTAR

Enhancing disruption prediction through Bayesian neural network in KSTAR

An advanced double-phase stacking ensemble technique with active learning classifier: Toward reliable disruption prediction in Aditya tokamak

A self-organised partition of the high dimensional plasma parameter space for plasma disruption prediction

Contact Info

Product

Resources

About