A Mathematical View of the EFIT-AI Project

The EFIT-AI project will create a modern advanced equilibrium reconstruction code capable of meeting the needs of tokamaks with burning plasmas. Mathematically, equilibrium reconstruction is an inverse problem, where one is seeking to use the data to infer underlying physical properties through the use of mathematical models for the forward problem. Inverse problems have enjoyed a renaissance of mathematical interest in recent years thanks to advances in theories in uncertainty quantification and machine learning. These advances are at the intersection of statistics, functional analysis, computer science, and physics. Here, we review some of these advances in the context of the equilibrium reconstruction problem.