Latent Space Mapping: Revolutionizing Predictive Models for Divertor Plasma Detachment Control

Ben Zhu, Menglong Zhao, Xue-Qiao Xu, Anchal Gupta, KyuBeen Kwon, Xinxing Ma, David Eldon · 2025

Read Latent Space Mapping: Revolutionizing Predictive Models for Divertor Plasma Detachment Control by Ben Zhu, Menglong Zhao, Xue-Qiao Xu, Anchal Gupta, KyuBeen Kwon, Xinxing Ma, David Eldon (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-19654.txt.

Description

Ben Zhu and co-authors built a fast stand in model in 2025 for the edge plasma of a fusion reactor, to help control a delicate state called detachment. Full simulations of that edge are far too slow for live control, so operators have relied on rough rules.

Their model learned from more than 70,000 detailed simulations of a Korean reactor. It compresses each plasma state into a compact hidden code and predicts from it in a fifth of a millisecond. That is a speedup of more than a hundred million times, with errors under twenty percent. In a 2024 test campaign, a control system using it managed detachment on its first try, even with a new wall material and no retuning.

Catalog record

CREATORBen Zhu, Menglong Zhao, Xue-Qiao Xu, Anchal Gupta, KyuBeen Kwon, Xinxing Ma, David Eldon
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE82 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2502-19654

Browse the full library · About the Tower of Babel