Towards Transparent and Accurate Plasma State Monitoring at JET

2025-02-14

AI Summary

Real-time, interpretable monitoring of plasma disruptions and other anomalous events in the JET tokamak remains challenging. Method: This work proposes a novel multi-task data-driven framework integrating supervised and unsupervised learning. It is the first to apply multi-task learning to plasma state monitoring, jointly modeling both normal and disruption evolutions. The framework incorporates physics-informed features (e.g., βₙ, q₉₅) and sequence modeling (LSTM/Transformer) to enhance robustness and interpretability of early disruption warnings. Latent space analysis further reveals separability between operational and disruption regimes and underlying dynamical patterns. Results: Experiments demonstrate excellent cross-validated accuracy and an average warning time of ~100 ms—sufficient for real-time intervention. The learned latent space distribution aligns with established plasma physics principles. The framework supports both disruption-triggered control actions and mechanistic investigation of disruption onset.

Technology Category

• Machine Learning: Multimodal Learning
• Intelligent Robots: State Estimation
• Computer Vision: Multi-modal Vision

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Controlling and monitoring plasma within a tokamak device is complex and challenging. Plasma off-normal events, such as disruptions, are hindering steady-state operation. For large devices, they can even endanger the machine's integrity and it represents in general one of the most serious concerns for the exploitation of the tokamak concept for future power plants. Effective plasma state monitoring carries the potential to enable an understanding of such phenomena and their evolution which is crucial for the successful operation of tokamaks. This paper presents the application of a transparent and data-driven methodology to monitor the plasma state in a tokamak. Compared to previous studies in the field, supervised and unsupervised learning techniques are combined. The dataset consisted of 520 expert-validated discharges from JET. The goal was to provide an interpretable plasma state representation for the JET operational space by leveraging multi-task learning for the first time in the context of plasma state monitoring. When evaluated as disruption predictors, a sequence-based approach showed significant improvements compared to the state-based models. The best resulting network achieved a promising cross-validated success rate when combined with a physical indicator and accounting for nearby instabilities. Qualitative evaluations of the learned latent space uncovered operational and disruptive regions as well as patterns related to learned dynamics and global feature importance. The applied methodology provides novel possibilities for the definition of triggers to switch between different control scenarios, data analysis, and learning as well as exploring latent dynamics for plasma state monitoring. It also showed promising quantitative and qualitative results with warning times suitable for avoidance purposes and distributions that are consistent with known physical mechanisms.

Problem

• Monitor plasma state in tokamak
• Predict plasma disruptions effectively
• Combine supervised and unsupervised learning

Innovation

• Combined supervised and unsupervised learning
• Multi-task learning for plasma monitoring
• Sequence-based disruption prediction approach