Saved in:
Bibliographic Details
Main Authors: Csala, Hunor, De Pascuale, Sebastian, Laiu, Paul, Lore, Jeremy, Park, Jae-Sun, Zhang, Pei
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2512.23884
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914225071325184
author Csala, Hunor
De Pascuale, Sebastian
Laiu, Paul
Lore, Jeremy
Park, Jae-Sun
Zhang, Pei
author_facet Csala, Hunor
De Pascuale, Sebastian
Laiu, Paul
Lore, Jeremy
Park, Jae-Sun
Zhang, Pei
contents Accurate modeling of scrape-off layer (SOL) and divertor-edge dynamics is vital for designing plasma-facing components in fusion devices. High-fidelity edge fluid/neutral codes such as SOLPS-ITER capture SOL physics with high accuracy, but their computational cost limits broad parameter scans and long transient studies. We present transformer-based, autoregressive surrogates for efficient prediction of 2D, time-dependent plasma edge state fields. Trained on SOLPS-ITER spatiotemporal data, the surrogates forecast electron temperature, electron density, and radiated power over extended horizons. We evaluate model variants trained with increasing autoregressive horizons (1-100 steps) on short- and long-horizon prediction tasks. Longer-horizon training systematically improves rollout stability and mitigates error accumulation, enabling stable predictions over hundreds to thousands of steps and reproducing key dynamical features such as the motion of high-radiation regions. Measured end-to-end wall-clock times show the surrogate is orders of magnitude faster than SOLPS-ITER, enabling rapid parameter exploration. Prediction accuracy degrades when the surrogate enters physical regimes not represented in the training dataset, motivating future work on data enrichment and physics-informed constraints. Overall, this approach provides a fast, accurate surrogate for computationally intensive plasma edge simulations, supporting rapid scenario exploration, control-oriented studies, and progress toward real-time applications in fusion devices.
format Preprint
id arxiv_https___arxiv_org_abs_2512_23884
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Autoregressive long-horizon prediction of plasma edge dynamics
Csala, Hunor
De Pascuale, Sebastian
Laiu, Paul
Lore, Jeremy
Park, Jae-Sun
Zhang, Pei
Plasma Physics
Artificial Intelligence
Accurate modeling of scrape-off layer (SOL) and divertor-edge dynamics is vital for designing plasma-facing components in fusion devices. High-fidelity edge fluid/neutral codes such as SOLPS-ITER capture SOL physics with high accuracy, but their computational cost limits broad parameter scans and long transient studies. We present transformer-based, autoregressive surrogates for efficient prediction of 2D, time-dependent plasma edge state fields. Trained on SOLPS-ITER spatiotemporal data, the surrogates forecast electron temperature, electron density, and radiated power over extended horizons. We evaluate model variants trained with increasing autoregressive horizons (1-100 steps) on short- and long-horizon prediction tasks. Longer-horizon training systematically improves rollout stability and mitigates error accumulation, enabling stable predictions over hundreds to thousands of steps and reproducing key dynamical features such as the motion of high-radiation regions. Measured end-to-end wall-clock times show the surrogate is orders of magnitude faster than SOLPS-ITER, enabling rapid parameter exploration. Prediction accuracy degrades when the surrogate enters physical regimes not represented in the training dataset, motivating future work on data enrichment and physics-informed constraints. Overall, this approach provides a fast, accurate surrogate for computationally intensive plasma edge simulations, supporting rapid scenario exploration, control-oriented studies, and progress toward real-time applications in fusion devices.
title Autoregressive long-horizon prediction of plasma edge dynamics
topic Plasma Physics
Artificial Intelligence
url https://arxiv.org/abs/2512.23884