Abstract:We present a data-driven reduced-order modeling of the space-charge dynamics for electromagnetic particle-in-cell (EMPIC) plasma simulations based on dynamic mode decomposition (DMD). The dynamics of the charged particles in kinetic plasma simulations such as EMPIC is manifested through the plasma current density defined along the edges of the spatial mesh. We showcase the efficacy of DMD in modeling the time evolution of current density through a low-dimensional feature space. Not only do such DMD-based predictive reduced-order models help accelerate EMPIC simulations, they also have the potential to facilitate investigative analysis and control applications. We demonstrate the proposed DMD-EMPIC scheme for reduced-order modeling of current density, and speed-up in EMPIC simulations involving electron beam under the influence of magnetic field, virtual cathode oscillations, and backward wave oscillator.

What problem does this paper attempt to address?

This paper attempts to solve the problem of high computational cost in electromagnetic particle - in - cell (EMPIC) simulations, especially the computational bottleneck caused by a large number of particles. Specifically, the paper proposes a data - driven reduced - order modeling method based on dynamic mode decomposition (DMD) for simulating space - charge dynamics, thereby accelerating EMPIC simulations. ### Background and Problem Description of the Paper **Background**: - **Electromagnetic Particle - in - Cell (EMPIC)**: Widely used in the modeling of high - power microwave sources, directed - energy devices, particle accelerators, terahertz devices, etc., as well as in the understanding of ionospheric phenomena, magnetospheric regions, and astrophysical events. - **Computational Challenges**: EMPIC simulations require a large amount of computational resources because the position and velocity of each super - particle need to be updated at each time step, covering the entire time window of interest. This leads to high computational costs, especially when dealing with a large number of particles. **Main Problems**: - **Computational Bottleneck**: In EMPIC simulations, due to the need to process a large number of particles, the computational cost is extremely high. At each time step, the position and velocity of each particle need to be updated, which is especially obvious in large - scale simulations. - **Reduced - Order Modeling Requirement**: In order to accelerate EMPIC simulations, it is necessary to develop reduced - order models (ROMs) that can capture the low - dimensional feature space of plasma dynamics to reduce the computational burden. ### Solutions **Methods**: - **Dynamic Mode Decomposition (DMD)**: DMD is a data - driven method for low - dimensional surrogate modeling of high - dimensional complex dynamical systems. The paper uses DMD to model the time evolution of current density, thereby avoiding the need to directly handle a large number of particles. - **DMD - EMPIC Algorithm**: - **Transient Phase (\( t \leq t_f \))**: Run a high - fidelity EMPIC simulation until the end of the transient, marked as the final time \( t_f \). At the same time, use the online DMD algorithm to detect the end of the transient in real - time. - **Post - Transient Phase (\( t > t_f \))**: After the end of the transient, use offline DMD to model snapshots of current density and predict future current density. Then, use the DMD - predicted current density to update the electromagnetic field without the need to handle particle motion anymore. **Main Contributions**: 1. **Constructed an interpretable reduced - order model for the first time**: By modeling current density with DMD, reduced - order modeling of space - charge dynamics in EMPIC simulations is achieved. This method not only accelerates EMPIC simulations but also helps in analyzing and diagnosing problems. 2. **Novel DMD - EMPIC Algorithm**: Demonstrated how to accelerate EMPIC simulations by predicting current density with DMD, especially for plasma systems showing long - time oscillation behavior (such as limit - cycle behavior). ### Results and Applications **Results Demonstration**: - **Example of Oscillating Electron Beam**: Through a 2D oscillating electron beam case, the effectiveness of the DMD - EMPIC algorithm is demonstrated. After the end of the transient, DMD can accurately predict the time evolution of current density, significantly reducing the computational time. **Application Prospects**: - **Accelerating Simulations**: The DMD - EMPIC algorithm can significantly accelerate EMPIC simulations, especially when dealing with large - scale particle systems. - **Control Theory Applications**: By modeling current density with DMD, a linear time - evolution model can be provided for control theory applications, further expanding its application range. In conclusion, this paper successfully solves the computational bottleneck problem in EMPIC simulations by introducing DMD technology, providing new tools and methods for plasma dynamics research.

Accelerating Particle-in-Cell Kinetic Plasma Simulations via Reduced-Order Modeling of Space-Charge Dynamics using Dynamic Mode Decomposition

Detection and prediction of equilibrium states in kinetic plasma simulations via mode tracking using reduced-order dynamic mode decomposition

Direct Implicit and Explicit Energy-Conserving Particle-in-Cell Methods for Modeling of Capacitively-Coupled Plasma Devices

Dynamic Mode Decomposition for data-driven analysis and reduced-order modelling of ExB plasmas: I. Extraction of spatiotemporally coherent patterns

Dynamic mode decomposition for plasma diagnostics and validation

Accelerating Kinetic Simulations of Electrostatic Plasmas with Reduced-Order Modeling

Dynamic Mode Decomposition for data-driven analysis and reduced-order modelling of ExB plasmas: II. dynamics forecasting

A 2-D Axisymmetric Magneto-Hydrodynamic Model of a DC Arc Plasma Torch and Its Solution Methodology

Kinetic simulations of low-pressure inductively coupled plasma: an implicit electromagnetic PIC/MCC model with the ADI-FDTD method

Three-Dimensional Direct-Implicit Particle-in-Cell Model Using Trilinear Anisotropic Immersed-Finite-Element for Plasma Propulsion

Magnetohydrodynamic with Adaptively Embedded Particle-in-Cell model: MHD-AEPIC

Geometric Particle-In-Cell discretizations of a plasma hybrid model with kinetic ions and mass-less fluid electrons

Data-driven discovery of reduced plasma physics models from fully kinetic simulations

Accuracy of the Explicit Energy-Conserving Particle-in-Cell Method for Under-resolved Simulations of Capacitively Coupled Plasma Discharges

Adaptive symplectic model order reduction of parametric particle-based Vlasov-Poisson equation

An Envelope Tracking Approach for Particle in Cell Simulations

High-order Implicit Particle-in-cell Method for Plasma Simulations at Solid Densities.

Study of scaling law for particle-in-cell/Monte Carlo simulation of low-temperature magnetized plasma for electric propulsion

An Unstructured Body-of-Revolution Electromagnetic Particle-in-Cell Algorithm with Radial Perfectly Matched Layers and Dual Polarizations

Generalized reduced-order particle-in-cell scheme for Hall thruster modeling: concept and in-depth verification in the axial-azimuthal configuration