Hybrid Fourier Neural Operator-Plasma Fluid Model for Fast and Accurate Multiscale Simulations of High Power Microwave Breakdown

• URL: http://arxiv.org/abs/2509.05799v1
• Date: Sat, 06 Sep 2025 18:24:33 GMT
• Title: Hybrid Fourier Neural Operator-Plasma Fluid Model for Fast and Accurate Multiscale Simulations of High Power Microwave Breakdown
• Authors: Kalp Pandya, Pratik Ghosh, Ajeya Mandikal, Shivam Gandha, Bhaskar Chaudhury,
• Abstract summary: We present a hybrid modeling approach that combines the accuracy of a differential equation-based plasma fluid solver with the computational efficiency of FNO.<n>Trained on data from an in-houseD-based plasma-fluid solver, the FNO replaces computationally expensive EM field updates.<n>Our work also demonstrate how such hybrid pipelines can be used to seamlessly integrate existing C-based simulation codes with Python-based machine learning frameworks.
• Score: 2.202064335120138
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Modeling and simulation of High Power Microwave (HPM) breakdown, a multiscale phenomenon, is computationally expensive and requires solving Maxwell's equations (EM solver) coupled with a plasma continuity equation (plasma solver). In this work, we present a hybrid modeling approach that combines the accuracy of a differential equation-based plasma fluid solver with the computational efficiency of FNO (Fourier Neural Operator) based EM solver. Trained on data from an in-house FDTD-based plasma-fluid solver, the FNO replaces computationally expensive EM field updates, while the plasma solver governs the dynamic plasma response. The hybrid model is validated on microwave streamer formation, due to diffusion ionization mechanism, in a 2D scenario for unseen incident electric fields corresponding to entirely new plasma streamer simulations not included in model training, showing excellent agreement with FDTD based fluid simulations in terms of streamer shape, velocity, and temporal evolution. This hybrid FNO based strategy delivers significant acceleration of the order of 60X compared to traditional simulations for the specified problem size and offers an efficient alternative for computationally demanding multiscale and multiphysics simulations involved in HPM breakdown. Our work also demonstrate how such hybrid pipelines can be used to seamlessly to integrate existing C-based simulation codes with Python-based machine learning frameworks for simulations of plasma science and engineering problems.

Related papers

• Case study of a differentiable heterogeneous multiphysics solver for a nuclear fusion application [39.74211852563065]
  This work presents a case study of a heterogeneous multiphysics solver from the nuclear fusion domain.<n>At the macroscopic scale, an auto-differentiable ODE solver in JAX computes the evolution of the pulsed power circuit and bulk plasma parameters for a compressing Z Pinch.<n>The ODE solver requires a closure for the impedance of the plasma load obtained via root-finding at every timestep, which we solve efficiently using gradient-based Newton iteration.
  arXiv  Detail & Related papers  (2025-11-17T11:23:16Z)
• Machine Learning-Integrated Hybrid Fluid-Kinetic Framework for Quantum Electrodynamic Laser Plasma Simulations [0.0]
  The research introduces a machine learning-based three-dimensional hybrid fluid-particle-in-cell (PIC) system.<n>The technique employs fluid approximations for stable areas but activates the PIC solver when SwitchNet directs it to unstable sections.<n>The model produces precise predictions with coefficient of determination (R2) values above 0.95 and mean squared errors below 10-4 for all field components.
  arXiv  Detail & Related papers  (2025-10-13T09:07:59Z)
• Hybrid Neural-MPM for Interactive Fluid Simulations in Real-Time [57.30651532625017]
  We present a novel hybrid method that integrates numerical simulation, neural physics, and generative control.<n>Our system demonstrates robust performance across diverse 2D/3D scenarios, material types, and obstacle interactions.<n>We promise to release both models and data upon acceptance.
  arXiv  Detail & Related papers  (2025-05-25T01:27:18Z)
• Particle-based plasma simulation using a graph neural network [0.0]
  The model achieves high accuracy with a time step longer than conventional simulation by two orders of magnitude.<n>This work demonstrates that complex plasma dynamics can be learned and shows promise for the development of fast differentiable simulators.
  arXiv  Detail & Related papers  (2025-03-01T01:07:37Z)
• Plasma Surrogate Modelling using Fourier Neural Operators [57.52074029826172]
  Predicting plasma evolution within a Tokamak reactor is crucial to realizing the goal of sustainable fusion. We demonstrate accurate predictions of evolution plasma using deep learning-based surrogate modelling tools, viz., Neural Operators (FNO) We show that FNO has a speedup of six orders of magnitude over traditional solvers in predicting the plasma dynamics simulated from magnetohydrodynamic models. FNOs can also predict plasma evolution on real-world experimental data observed by the cameras positioned within the MAST Tokamak.
  arXiv  Detail & Related papers  (2023-11-10T10:05:00Z)
• Hybridizing Physics and Neural ODEs for Predicting Plasma Inductance Dynamics in Tokamak Fusion Reactors [0.0]
  We train both physics-based and neural network models on data from the Alcator C-Mod fusion reactor. We find that a model that combines physics-based equations with a neural ODE performs better than both existing physics-motivated ODEs and a pure neural ODE model.
  arXiv  Detail & Related papers  (2023-10-30T23:25:54Z)
• Machine learning of hidden variables in multiscale fluid simulation [77.34726150561087]
  Solving fluid dynamics equations often requires the use of closure relations that account for missing microphysics. In our study, a partial differential equation simulator that is end-to-end differentiable is used to train judiciously placed neural networks. We show that this method enables an equation based approach to reproduce non-linear, large Knudsen number plasma physics.
  arXiv  Detail & Related papers  (2023-06-19T06:02:53Z)
• Unsupervised Discovery of Inertial-Fusion Plasma Physics using Differentiable Kinetic Simulations and a Maximum Entropy Loss Function [77.34726150561087]
  We create a differentiable solver for the plasma kinetics 3D partial-differential-equation and introduce a domain-specific objective function. We apply this framework to an inertial-fusion relevant configuration and find that the optimization process exploits a novel physical effect.
  arXiv  Detail & Related papers  (2022-06-03T15:27:33Z)
• Deep Learning Architecture Based Approach For 2D-Simulation of Microwave Plasma Interaction [5.467400475482668]
  This paper presents a convolutional neural network (CNN)-based deep learning model, inspired from UNet with series of encoder and decoder units with skip connections, for the simulation of microwave-plasma interaction. The training data associated with microwave-plasma interaction has been generated using 2D-FDTD (Finite Difference Time Domain) based simulations. The trained deep learning model is then used to reproduce the scattered electric field values for the 1GHz incident microwave on different plasma profiles with error margin of less than 2%.
  arXiv  Detail & Related papers  (2022-06-02T19:37:41Z)
• Using neural networks to solve the 2D Poisson equation for electric field computation in plasma fluid simulations [0.0]
  The Poisson equation is critical to get a self-consistent solution in plasma fluid simulations used for Hall effect thrusters and streamers discharges. solving the 2D Poisson equation with zero Dirichlet boundary conditions using a deep neural network is investigated. A CNN is built to solve the same Poisson equation but in cylindrical coordinates. Results reveal good CNN predictions with significant speedup.
  arXiv  Detail & Related papers  (2021-09-27T14:25:10Z)
• Hybridized Methods for Quantum Simulation in the Interaction Picture [69.02115180674885]
  We provide a framework that allows different simulation methods to be hybridized and thereby improve performance for interaction picture simulations. Physical applications of these hybridized methods yield a gate complexity scaling as $log2 Lambda$ in the electric cutoff. For the general problem of Hamiltonian simulation subject to dynamical constraints, these methods yield a query complexity independent of the penalty parameter $lambda$ used to impose an energy cost.
  arXiv  Detail & Related papers  (2021-09-07T20:01:22Z)
• Combining Differentiable PDE Solvers and Graph Neural Networks for Fluid Flow Prediction [79.81193813215872]
  We develop a hybrid (graph) neural network that combines a traditional graph convolutional network with an embedded differentiable fluid dynamics simulator inside the network itself. We show that we can both generalize well to new situations and benefit from the substantial speedup of neural network CFD predictions.
  arXiv  Detail & Related papers  (2020-07-08T21:23:19Z)
• A Comparative Study of Machine Learning Models for Predicting the State of Reactive Mixing [0.0]
  Accurate predictions of reactive mixing are critical for many Earth and environmental science problems. A high-fidelity, finite-element-based numerical model is built to solve the fast, irreversible bimolecular reaction-diffusion scenarios. A total of 2,315 simulations are performed using different sets of model input parameters.
  arXiv  Detail & Related papers  (2020-02-24T22:50:19Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.