Machine Learning Design of Soft Magnetoresponsive Materials With Multidirectional Fields
SmartMat · 2025-12 · 6 (6)
DOI: 10.1002/smm2.70051 ↗Kurz gefasst
Fachartikel in der Zeitschrift SmartMat, online erschienen am 18.11.2025. Verfasst von Xin Ye und 7 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 58 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Soft magnetoresponsive materials (SMRMs) with programmable magnetization profiles can exhibit a wide range of motion postures under dynamically controllable magnetic fields. Although the complex magnetization distribution provides a vast design space, the intricate nonlinear deformation and magneto‐mechanical coupling mechanisms make the design process highly challenging. The prevailing design methodologies are limited by their inefficiency and a lack of precision, which constrain the application potential of SMRMs. Here, we propose a method that combines machine learning and evolutionary algorithms to achieve forward prediction of deformations of SMRMs and inverse design of the magnetization profiles required for target deformations. A machine learning model based on a fully connected neural network (FCNN) is trained to predict deformations under external magnetic fields from six directions with high accuracy, within 0.3 ms. Based on this, an evolutionary algorithm is introduced to search for the inverse design solutions that achieve the target deformations from a pool of millions of candidates within minutes. By incorporating two symmetries into the physical model, the required amount of training data is reduced by 75%, enabling the design of controllable dynamic fluctuations and multi‐modal arrays. An image recognition method has been developed to transform natural curves into deformation targets, demonstrating the static reconstruction of leaf contours and the dynamic replication of four distinct fish swimming modes. This approach provides an efficient design tool and expands the application scope of the material.
Zitieren
APA
Ye, X., Zhang, Y., Zhu, H., Huang, H., Lv, R., Zhao, S., Zhao, Y., & Jiang, C. (2025). Machine Learning Design of Soft Magnetoresponsive Materials With Multidirectional Fields. SmartMat, 6(6). https://doi.org/10.1002/smm2.70051
Vancouver
Ye X, Zhang Y, Zhu H, Huang H, Lv R, Zhao S, et al. Machine Learning Design of Soft Magnetoresponsive Materials With Multidirectional Fields. SmartMat. 2025;6(6). doi:10.1002/smm2.70051
BibTeX
@article{Ye2025,
author = {Ye, Xin and Zhang, Yilong and Zhu, Hanlin and Huang, Hui and Lv, Rui and Zhao, Shiwei and Zhao, Yan and Jiang, Chao},
title = {Machine Learning Design of Soft Magnetoresponsive Materials With Multidirectional Fields},
journal = {SmartMat},
year = {2025},
volume = {6},
number = {6},
doi = {10.1002/smm2.70051}
}
EUVIMED