The use of bayesian inversion to resolve plasma equilibrium

Area of Science:

• Plasma physics
• Computational physics
• Fusion energy research

Background:

• Bayesian probability theory enhances accuracy in analyzing complex experimental data.
• Probabilistic frameworks like MINERVA integrate uncertainties and prior knowledge for improved physics variable inference.
• Existing methods often rely on explicit equilibrium assumptions, limiting flexibility.

Purpose of the Study:

• Develop advanced Bayesian inversion tools for plasma physics.
• Distinguish between competing equilibrium theories using data from the MAST spherical tokamak.
• Test magnetohydrodynamics (MHD) theory predictions, specifically mode structure, on the H-1 Heliac.

Main Methods:

• Utilized a Bayesian graphical model framework (MINERVA) for data analysis.
• Performed Bayesian inversion for poloidal magnetic flux without explicit equilibrium assumptions.
• Integrated data from motional Stark effect, pickup coils, and flux loops for constrained equilibrium inference.

Main Results:

• Successfully applied Bayesian inversion to infer internal magnetic structure.
• Demonstrated the capability to distinguish between different plasma equilibrium theories.
• Validated MHD theory predictions regarding mode structure in fusion devices.

Conclusions:

• Bayesian inversion offers a powerful, flexible approach to plasma physics analysis.
• The developed tools enhance the accuracy of inferred plasma variables and theoretical testing.
• This work advances the understanding of plasma behavior in fusion devices.