Visible-Light Imaging Diagnosis of Neutral Particle Emission Tomography in the Tokamak Divertor: An Efficient Transformer-based Surrogate Model

2026-07-19

AI Summary

This work addresses the challenge of efficiently reconstructing the two-dimensional emission distribution of neutral particles in tokamak divertors by proposing Delta-InvFormer, a lightweight encoder–decoder network based on a differential self-attention mechanism. Leveraging spatiotemporal images captured by visible-light cameras, the model employs a differential Transformer architecture to effectively suppress noise and integrate dynamic features, thereby enabling high-fidelity prediction of light intensity distributions. Experimental results on real data from the EAST tokamak demonstrate that Delta-InvFormer achieves competitive reconstruction accuracy while significantly accelerating inference compared to conventional approaches, offering an efficient and real-time alternative for fusion plasma diagnostics.

Abstract

Nuclear fusion has made significant progress in recent years and is expected to become one of the most important pathways to addressing global energy challenges. This paper focuses on observing plasma using visible-light cameras, analyzing its spatio-temporal motion cues, and predicting the two-dimensional spatial distribution of light intensity, aiming to provide a foundational basis for future scientific experiments using deep neural networks. Specifically, we propose Delta-InvFormer, a novel backbone network centered on a differential Transformer. The key insight is that by taking consecutive video frames as input, we can better capture the dynamics of the plasma. Moreover, spatial and temporal differential self-attention effectively mitigates interference from noisy signals, ensuring high-quality feature extraction. These features are then fused into a compact and informative representation, which is fed into a decoder network to predict the distribution. Based on real experimental data collected from the Experimental Advanced Superconducting Tokamak (EAST) large-scale scientific facility, our results demonstrate that the proposed model not only significantly accelerates traditional methods for distribution prediction but also achieves competitive reconstruction accuracy. The source code of this paper will be released on https://github.com/Event-AHU/OpenFusion

Problem

• Visible-Light Imaging
• Neutral Particle Emission Tomography
• Tokamak Divertor
• Plasma Diagnostics
• Light Intensity Distribution

Innovation

• Differential Transformer
• Surrogate Model
• Plasma Diagnostics
• Visible-Light Imaging
• Neutral Particle Emission Tomography