Autoregressive long-horizon prediction of plasma edge dynamics
Preprint
Abstract
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.Keywords
All Related Versions
This publication has 15 references indexed in Scilit:
- Effect of time step, neutral–neutral collisions, and an underrelaxation scheme on the numerical convergence of SOLPS-ITER plasma boundary simulations with kinetic neutralsPhysics of Plasmas, 2025
- Neural operator surrogate models of plasma edge simulations: feasibility and data efficiencyNuclear Fusion, 2025
- Latent space mapping: Revolutionizing predictive models for divertor plasma detachment controlPhysics of Plasmas, 2025
- Multi-Output Prediction of HL-2A Device Boundary Characteristic Quantities Based on Machine LearningJournal of Fusion Energy, 2025
- Full time-dependent SOLPS-ITER simulation of the SPARC tokamak: actuator design for particle and divertor condition control *Nuclear Fusion, 2024
- Fast dynamic 1D simulation of divertor plasmas with neural PDE surrogatesNuclear Fusion, 2023
- Bifurcation-like transition of divertor conditions induced by X-point radiation in KSTAR L-mode plasmas *Nuclear Fusion, 2023
- Towards fast surrogate models for interpolation of tokamak edge plasmasNuclear Materials and Energy, 2023
- Benchmark of a self-consistent dynamic 1D divertor model DIV1D using the 2D SOLPS-ITER codePlasma Physics and Controlled Fusion, 2022
- Data-driven model for divertor plasma detachment predictionJournal of Plasma Physics, 2022