This work explores the application of deep operator learning principles to a problem in statistical physics. Specifically, we consider the linear kinetic equation, consisting of a differential advection operator and an integral collision operator, which is a powerful yet expensive mathematical model for interacting particle systems with ample applications, e.g., in radiation transport. We investigate the capabilities of the Deep Operator network (DeepONet) approach to modelling the high dimensional collision operator of the linear kinetic equation. This integral operator has crucial analytical structures that a surrogate model, e.g., a DeepONet, needs to preserve to enable meaningful physical simulation. We propose several DeepONet modifications to encapsulate essential structural properties of this integral operator in a DeepONet model. To be precise, we adapt the architecture of the trunk-net so the DeepONet has the same collision invariants as the theoretical kinetic collision operator, thus preserving conserved quantities, e.g., mass, of the modeled many-particle system. Further, we propose an entropy-inspired data-sampling method tailored to train the modified DeepONet surrogates without requiring an excessive expensive simulation-based data generation.
First-order continuous- and discontinuous-Galerkin moment models for a linear kinetic equation: realizability-preserving splitting scheme and numerical analysis
A realizability-preserving high-order kinetic scheme using WENO reconstruction for entropy-based moment closures of linear kinetic equations in slab geometry
TU Kaiserslautern（凯泽斯劳滕工业大学）；RWTH Aachen University（亚琛工业大学）
Asymptotic preserving schemes for highly oscillatory kinetic equation
INRIA-Rennes Bretagne Atlantique（法国国家信息与自动化研究所雷恩布列塔尼大西洋分中心）；CNRS and IRMAR, Université de Rennes 1（法国国家科学研究中心与雷恩第一大学IRMAR）；IRMAR, Université de Rennes 1（雷恩第一大学IRMAR）
Well-balanced and asymptotic preserving schemes for kinetic models
Sorbonne Universités, UPMC Univ Paris 06（索邦大学，巴黎第六大学）；CNRS（法国国家科学研究中心）；INRIA-Paris-Rocquencourt（法国国家信息与自动化研究所巴黎-罗克安库尔分部）；Institute of natural sciences and department of mathematics, Shanghai Jiao Tong University（上海交通大学自然科学研究院和数学系）
Hypocoercivity based Sensitivity Analysis and Spectral Convergence of the Stochastic Galerkin Approximation to Collisional Kinetic Equations with Multiple Scales and Random Inputs
The Institute for Computational Engineering and Sciences (ICES), University of Texas at Austin（德克萨斯大学奥斯汀分校计算工程与科学研究所）；Department of Mathematics, University of Wisconsin-Madison（威斯康星大学麦迪逊分校数学系）
Asymptotic preserving Implicit-Explicit Runge-Kutta methods for non linear kinetic equations
Université de Toulouse（图卢兹大学；法国国家科学研究中心数学研究所）；UPS, INSA, UT1, UTM（费拉拉大学数学系与CMCS）；CNRS, UMR 5219；Institut de Mathématiques de Toulouse；Mathematics Department, University of Ferrara and CMCS
Structure-preserving marker-particle discretizations of Coulomb collisions for particle-in-cell codes
本文为等离子体动力学模型中库仑碰撞的离散化提供了新的见解。基于以前的工作[Carrillo et al. J. Comp. Phys. X 7：100066（2020），Hirvijoki and Burby Phys. Plasmas 27（8）：082307（2020）]，我提出了标准Landau碰撞算子和静电回旋Landau算子的确定性离散时间能量和正性保持，熵耗散标记粒子方案。在标准朗道算子的情况下，该方案还保留了离散时间的动力学动量。这些改进，[Carrillo et al. J. Comp. Phys. X 7：100066（2020），Hirvijoki and Burby Phys. Plasmas 27（8）：082307（2020）]中的结构保持离散化到离散时间的扩展，通过利用所涉及的碰撞算子和所谓的离散梯度积分器的底层metriplectic结构而成为可能。
