• Corpus ID: 246275786

Neural network-based, structure-preserving entropy closures for the Boltzmann moment system

  title={Neural network-based, structure-preserving entropy closures for the Boltzmann moment system},
  author={Steffen Schotth{\"o}fer and Tianbai Xiao and Martin Frank and Cory D. Hauck},
This work presents neural network based minimal entropy closures for the moment system of the Boltzmann equation, that preserve the inherent structure of the system of partial differential equations, such as entropy dissipation and hyperbolicity. The described method embeds convexity of the moment to entropy map in the neural network approximation to preserve the structure of the minimal entropy closure. Two techniques are used to implement the methods. The first approach approximates the map… 
Predicting continuum breakdown with deep neural networks
The multi-scale nature of gaseous flows poses tremendous difficulties for theoretical and numerical analysis. The Boltzmann equation, while possessing a wider applicability than hydrodynamic
KiT-RT: An extendable framework for radiative transfer and therapy
KiT-RT (Kinetic Transport Solver for Radiation Therapy), an open-source C++ based framework for solving kinetic equations in radiation therapy applications is presented, to provide a collection of classical deterministic solvers for unstructured meshes that allow for easy extendability.


Data-driven, structure-preserving approximations to entropy-based moment closures for kinetic equations
A data-driven approach to construct entropy-based closures for the moment system from kinetic equations that provide accurate solutions in much less computation time than the standard, optimization-based MN closures.
A Regularized Entropy-Based Moment Method for Kinetic Equations
It is suggested that with proper choice of the regularization parameter, the new method can be used to generate accurate solutions of the original entropy-based moment method, and this is confirmed with numerical simulations.
High-Order Entropy-Based Closures for Linear Transport in Slab Geometry II: A Computational Study of the Optimization Problem
A numerical algorithm to implement entropy-based moment models in the context of a simple, linear kinetic equation for particles moving through a material slab and a Newton-based solver for the dual optimization problem, which uses an adaptive quadrature to evaluate integrals in the dual objective and its derivatives.
Using neural networks to accelerate the solution of the Boltzmann equation
Convex duality and entropy-based moment closures: Characterizing degenerate densities
It is shown that the set of degenerate densities is a union of convex cones defined by the complementary slackness conditions and, under reasonable assumptions, that this set is small in both a topological and a measure-theoretic sense.
Machine learning moment closure models for the radiative transfer equation III: enforcing hyperbolicity and physical characteristic speeds
A new method to enforce the hyperbolicity of the ML closure model is proposed and two new neural network architectures are designed based on the observation that the coefficient matrix of the closure system is a lower Hessenberg matrix.
Optimization and large scale computation of an entropy-based moment closure
Machine learning surrogate models for Landau fluid closure
It is found that in order to train a well-preformed network, a minimum size of the training data set is needed; MLP requires a minimum number of neurons in the hidden layers that equals the degrees of freedom in Fourier space, despite the fact that training data are being fed into the configuration space.
Machine learning based non-Newtonian fluid model with molecular fidelity
A machine-learning-based framework for constructing a non-Newtonian fluid dynamics model directly from a microscale description, with a generalized form of the objective tensor derivative that retains the microscale interpretations.