https://fusion.zoom.us/j/92881124411
Abstract: The characterization of fast-ion populations in magnetically confined fusion plasmas is crucial for understanding energy transfer mechanisms and optimizing plasma performance in fusion reactors. Fast ions, generated through nuclear reactions and auxiliary heating methods, play a pivotal role in the plasma heating process and stability. This seminar presents an analysis of fast-ion loss detector (FILD) measurements aimed at determining the fast-ion velocity distribution function. The discussion extends to the mathematical and computational challenges associated with extracting this velocity distribution from FILD data, highlighting the nature of these challenges as ill-posed inverse problems. These problems, where solutions may not exist, be non-unique, or not depend continuously on the data, require sophisticated solution techniques to obtain physically meaningful insights. The seminar explores the application of Tikhonov regularization, iterative reconstruction techniques, anisotropic regularization, and deep learning approaches to stabilize and solve these ill-posed problems. Each method offers a unique balance between fidelity to the experimental data and the imposition of prior information. The aim is to offer a comprehensive overview of the methodologies employed to decipher the fast-ion dynamics in fusion plasmas, presenting recent results that underscore the potential of these techniques to enhance understanding of fast-ion behavior.