NICFD.rst

.. sectionauthor:: Evert Bunschoten

Machine learning methods for NICFD

Important

Documentation in development

Process Diagram

# TODO: process diagram

# TODO: describe process on training EEoS-PINN networks from scratch

Pre-Processing

# TODO: distinction between state variables and train variables

# TODO: Setters for scaling methods: distinction between state variables and training variable(s)

Physics-Informed Training Loss Function

# TODO: Describe entropy-based equation of state

# TODO: Jacobi-based loss function

Training Progress and Output:

# TODO: describe which files are saved: thermodynamic state data plots, training loss functions for state variables.