 |
General: 663 |
P1: Modeling Turbulence in Tokamak Plasmas using Reservoir Computing
Author: Nathaniel Barbour , Advisor: William Dorland (Physics Department)
Problem Statement Presentation
Abstract
Turbulence in tokamaks, a geometric configuration for magnetic plasma fusion experiments, is a significant catalyst for driving the radial transport of heat. To increase the efficiency of magnetic fusion experiments, radial heat flux must be mitigated. When radial heat flux increases, additional energy must be introduced into the system via external heating in order to sustain the core temperatures required for fusion reactions to occur. Our objective for AMSC 663-664 is to leverage recent advances in machine learning approaches to the modeling and prediction of properties of complex systems to study turbulence in tokamak plasmas. We aim to develop a machine-learning-based tool that can receive state parameters of the plasma and the equations governing the time evolution of the system as inputs and can both accurately and quickly approximate the relevant macroscopic variables necessary to calculate the radial heat flux and other salient physics properties.
Final Presentation ,
Final Report
