Parameter Identification for Magneto‐Mechanical Resonators by a Convolutional Neural Network
ABSTRACT Magneto‐mechanical resonators (MMRs) are a promising new type of passive sensor. Their use requires the real‐time solution of a parameter identification problem where frequency and damping of an oscillatory motion have to be inferred from a time sequence of deflection angles. We investigate the performance of a simple convolutional neural network (CNN), trained solely on synthetic data. Our results demonstrate that the CNN can give fast reconstructions of the natural frequency in very little time but struggles to robustly estimate the rate of damping.
Authors
- Daniel Ruprecht (ORCID: https://orcid.org/0000-0003-1904-2473)
- Tobias A. Knopp (ORCID: https://orcid.org/0000-0002-1589-8517)
- Sebastian Götschel (ORCID: https://orcid.org/0000-0003-0287-2120)
- Martin Möddel (ORCID: https://orcid.org/0000-0002-4737-7863)
- Sarah Reiss (ORCID: https://orcid.org/0009-0006-4015-1200)
- UFUK AKKAYA (ORCID: https://orcid.org/0000-0003-1817-4255)
- Finn Sommer (ORCID: https://orcid.org/0009-0008-6977-9725)
Institutions
- Universität Hamburg (DE)
- University Medical Center Hamburg-Eppendorf (DE)
- Fraunhofer-Einrichtung für Individualisierte Medizintechnik (DE)
- Hamburg University of Technology (DE)
Publication Details
- Journal
- PAMM
- Published
- 2026-11-16
- DOI
- https://doi.org/10.1002/pamm.70270
- Primary Topic
- Mechanical and Optical Resonators
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Deutsche Forschungsgemeinschaft