EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Multiscale and Multi‐Timestep Switching of Multiple Machine Learning Force Fields for Artificial Intelligence‐Driven Materials Simulations

Ryuya Kanda, Megumu Yamazaki, Yuta Yoshimoto, Shintaro Izumi, Atsuki Inoue, Hiroshi Kawaguchi, Yasufumi Sakai

Advanced Intelligent Discovery · 04.08.2026

DOI: 10.1002/aidi.70152 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Advanced Intelligent Discovery, online erschienen am 04.08.2026. Verfasst von Ryuya Kanda und 6 weiteren Autor:innen. Die Arbeit nennt 21 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Molecular dynamics (MD) is essential for investigating atomic‐scale processes in materials and molecular systems, but the cost of high‐accuracy machine learning force field simulations still limits accessible system sizes and timescales. Here, we propose a practical model‐switching strategy for Deep Potential (DP)‐based MD simulations that alternates between independently trained DP models with different cutoff radii: a standard 6 Å model for higher accuracy and a reduced‐cutoff 4 Å model for faster inference. The method was implemented in LAMMPS/DeePMD and evaluated using solid‐phase anatase TiO 2 and liquid‐phase polyethylene glycol (PEG). For anatase TiO 2 , the 1:3 4–6 Å switching scheme preserved radial distribution function (RDF) correlations of 0.996 or higher relative to the 6 Å baseline while achieving a 1.24‐fold speedup. For PEG, the switching scheme maintained RDF correlations of 0.996 or higher with a 1.18‐fold speedup. Additional optimization using network‐size reduction and mixed‐precision inference achieved a 2.53‐fold speedup with RDF correlations of 0.975–0.988. Constant particle‐number, pressure, and temperature (NPT) simulations remained stable, whereas constant particle‐number, volume, and energy (NVE) simulations revealed system‐dependent energy‐drift behavior, particularly for aggressively optimized models. These results demonstrate that DP model switching provides a simple and practical route for accelerating structural MD simulations while highlighting the need for validation when strict energy conservation is required.

Zitieren

APA

Kanda, R., Yamazaki, M., Yoshimoto, Y., Izumi, S., Inoue, A., Kawaguchi, H., & Sakai, Y. (2026). Multiscale and Multi‐Timestep Switching of Multiple Machine Learning Force Fields for Artificial Intelligence‐Driven Materials Simulations. Advanced Intelligent Discovery. https://doi.org/10.1002/aidi.70152

Vancouver

Kanda R, Yamazaki M, Yoshimoto Y, Izumi S, Inoue A, Kawaguchi H, et al. Multiscale and Multi‐Timestep Switching of Multiple Machine Learning Force Fields for Artificial Intelligence‐Driven Materials Simulations. Advanced Intelligent Discovery. 2026. doi:10.1002/aidi.70152

BibTeX
@article{Kanda2026,
  author = {Kanda, Ryuya and Yamazaki, Megumu and Yoshimoto, Yuta and Izumi, Shintaro and Inoue, Atsuki and Kawaguchi, Hiroshi and Sakai, Yasufumi},
  title = {Multiscale and Multi‐Timestep Switching of Multiple Machine Learning Force Fields for Artificial Intelligence‐Driven Materials Simulations},
  journal = {Advanced Intelligent Discovery},
  year = {2026},
  doi = {10.1002/aidi.70152}
}
Kostenlos starten →