TGLF-SINN: Deep Learning Surrogate Model for Accelerating Turbulent Transport Modeling in Fusion

2025-09-07

AI Summary

In tokamak whole-device turbulent transport simulations, the TGLF model suffers from high computational cost, while existing neural-network-based surrogate models require prohibitively large training datasets. To address these challenges, we propose the Spectral Information Neural Network (SINN): a physics-informed surrogate that extracts dominant transport-spectrum patterns via physics-guided feature engineering, enforces physical consistency through transport-spectrum regularization, and employs Bayesian active learning to efficiently select the most informative training samples. SINN achieves near-full-dataset accuracy using only 25% of the training data, accelerates inference by 45×, and significantly reduces the normalized root-mean-square error (LRMSE) compared to baseline surrogates. By integrating differentiability, high fidelity, computational efficiency, and strong generalizability, SINN provides a practical, scalable acceleration framework for large-scale integrated tokamak simulations.

Technology Category

• Machine Learning: Active Learning
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

The Trapped Gyro-Landau Fluid (TGLF) model provides fast, accurate predictions of turbulent transport in tokamaks, but whole device simulations requiring thousands of evaluations remain computationally expensive. Neural network (NN) surrogates offer accelerated inference with fully differentiable approximations that enable gradient-based coupling but typically require large training datasets to capture transport flux variations across plasma conditions, creating significant training burden and limiting applicability to expensive gyrokinetic simulations. We propose extbf{TGLF-SINN (Spectra-Informed Neural Network)} with three key innovations: (1) principled feature engineering that reduces target prediction range, simplifying the learning task; (2) physics-guided regularization of transport spectra to improve generalization under sparse data; and (3) Bayesian Active Learning (BAL) to strategically select training samples based on model uncertainty, reducing data requirements while maintaining accuracy. Our approach achieves superior performance with significantly less training data. In offline settings, TGLF-SINN reduces logarithmic root mean squared error (LRMSE) by 12. 4% compared to the current baseline ase. Using only 25% of the complete dataset with BAL, we achieve LRMSE only 0.0165 higher than ase~and 0.0248 higher than our offline model (0.0583). In downstream flux matching applications, our NN surrogate provides 45x speedup over TGLF while maintaining comparable accuracy, demonstrating potential for training efficient surrogates for higher-fidelity models where data acquisition is costly and sparse.

Problem

• Accelerating turbulent transport modeling in fusion devices
• Reducing computational cost of whole device simulations
• Minimizing training data requirements for neural surrogates

Innovation

• Principled feature engineering reduces prediction range complexity
• Physics-guided regularization improves sparse data generalization
• Bayesian Active Learning strategically selects training samples