Year: 2025
Type: article
Abstract: The design of fusion devices is typically based on computationally expensive simulations. This can be alleviated using high aspect ratio models that employ a reduced number of free parameters, especially in the case of stellarator optimization where non-axisymmetric magnetic fields with a large parameter space are optimized to satisfy certain performance criteria. However, optimization is still required to find configurations with properties such as low elongation, high rotational transform, finite beta and good fast particle confinement. In this work, we train a machine learning model to construct configurations with favourable confinement properties by finding a solution to the inverse design problem, that is, obtaining a set of model input parameters for given desired properties. Since the solution of the inverse problem is non-unique, a probabilistic approach, based on mixture density networks, is used. It is shown that optimized configurations can be generated reliably using this method.
Source: Journal of Plasma Physics
Authors: P. Curvo, D.R. Ferreira, R. Jorge
Language: English
Subfield: Nuclear and High Energy Physics
Field: Physics and Astronomy
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid: $3,550
Funders: National Science Foundation, U.S. Department of Energy, National Energy Research Scientific Computing Center, EUROfusion, European Commission, Office of Science, Euratom Research and Training Programme, Fundação para a Ciência e a Tecnologia, Office of Advanced Cyberinfrastructure, Oak Ridge National Laboratory
Awards: G115085052, G121915436, G122673668, G1286236842, CDS&E: Moment Approach to Multiscale Plasmas, G1453181110, G1645119126, G1677143136, G1863003419, Track 4: Advanced CI Coordination Ecosystem: Monitoring and Measurement Services, G203625172, Implementation of activities described in the Roadmap to Fusion during Horizon Europe through a joint programme of the members of the EUROfusion consortium, G2293481551, G2411683797, G2988778103, G3083819904, G3510889102, G3562386289, G3562821882, G400305826, G419881185, G42916606, G4423657506, G4499487552, G4501827968, ACO: An Open CI Ecosystem to Advance Scientific Discovery (OpenCI), G498139845, G5076365615, Track 3: COre National Ecosystem for CyberinfrasTructure (CONECT), G5172252545, Instituto de Plasmas e Fusão Nuclear, G526230902, Instituto de Plasmas e Fusão Nuclear, G5367127336, G5464919783, Instituto de Plasmas e Fusão Nuclear, G660579501, G675320460, G6997658380, G7651947074, G8103039852, G8269158468, Track 2: Customized Multi-tier Assistance, Training, and Computational Help (MATCH) for End User ACCESS to CI, G8862356818, G8906985441, G969889393
OpenAlex ID: W4407677618