Simultaneous inference of multiple plasma parameter profiles by utilizing transport properties

Abstract

Numerical techniques for the computation of posterior distribution ease restrictions on inference models, and multiple plasma parameters can be simultaneously inferred with detailed modelling of uncertainties. In addition, data that depend on multiple parameters in a complex manner can also be utilized. This paper discusses frameworks that further expand the inference capability by including transport properties in plasmas. A prior belief based on a transport equation helps to localize the posterior distribution, and when ample diagnostic capability is available, even unknown parameters in the transport model can be estimated. Furthermore, these transport parameters can be directly inferred when the profiles of the measurable plasma parameters can be easily calculated for the given transport properties.

Special issue: Machine learning methods in plasma physics

“…The recent trend in machine learning, which incorporates physics knowledge, is reflected in the papers of T. Nishizawa [ 13 ] and T. M. Tyranowski et al [ 14 ] In the former paper, transport parameters are inferred based on an integrated data analysis approach, using an appropriately constrained Gaussian process. In the latter paper, a reduced complexity model for the kinetic Vlasov equation is derived, taking the underlying Hamiltonian structure of the Vlasov equation into account.…”