Temperature Field Reconstruction of Tungsten Monoblock Divertor on EAST using Physics-aware Neural Operator Transformer

2026-06-30

AI Summary

This work addresses the challenge of efficiently reconstructing the temperature field in tokamak divertors—a task critical for real-time plasma control yet poorly suited to traditional finite element methods. To overcome this limitation, the authors propose a Physics-aware Neural Operator Transformer (PNOT) that integrates graph attention mechanisms to model boundary heat flux relationships, embeds physical constraint operators, and employs Sobolev regularization in the loss function. By synergistically combining neural operators, physics-informed deep learning, and Transformer architecture, PNOT ensures physically consistent solutions while achieving high-fidelity spatiotemporal reconstruction. The method demonstrates markedly improved prediction accuracy and generalization capability, meeting the stringent demands of real-time applications in fusion energy systems.

Abstract

Accurate modeling of the divertor temperature field is essential for preventing material melting and damage and for extending the service life of fusion devices. However, conventional numerical methods, such as the Finite Element Method (FEM), are computationally expensive and therefore unsuitable for real-time applications. Therefore, a fast and generalizable method is required for real-time reconstruction of the divertor temperature field and subsequent real-time control. To address the above issue, we propose a Physics-aware Neural Operator Transformer (PNOT) to characterize the spatiotemporal evolution of the divertor temperature field. It models boundary heat-flux relations as a structured graph and employs graph attention to explicitly capture spatial physical dependencies. Inspired by physics-aware attention, we further develop a physics-aware neural operator module to aggregate query points with similar physical conditions via slicing and model heat diffusion, while a gradient-constrained Sobolev regularization loss enforces consistency between function values and their derivatives. Experimental results show that these physical constraints improve prediction accuracy while preserving physical consistency. The source code of this paper will be released on https://github.com/Event-AHU/OpenFusion

Problem

• temperature field reconstruction
• divertor
• real-time modeling
• fusion devices
• computational efficiency

Innovation

• Physics-aware Neural Operator
• Graph Attention
• Sobolev Regularization
• Temperature Field Reconstruction
• Real-time Fusion Control