| Title: |
Ultrasound-based Condition Monitoring of Power Converters with Physics-Informed Compression |
| Authors: |
Fassi, Youssof; Heiries, Vincent; Boutet, J.; Marianne, Julien; Martin, Sébastien; Chareyron, Mathilde; Chambon, Clément; Boisseau, Sébastien |
| Contributors: |
Commissariat à l'énergie atomique et aux énergies alternatives - Laboratoire d'Electronique et de Technologie de l'Information (CEA-LETI); Direction de Recherche Technologique (CEA) (DRT (CEA)); Commissariat à l'énergie atomique et aux énergies alternatives (CEA)-Commissariat à l'énergie atomique et aux énergies alternatives (CEA); SERMA INGENIERIE; This work is part of the IPCEI Microelectronicsand Connectivity and was supported by theFrench Public Authorities within the frame ofFrance 2030.; GDR SEEDS France & EPE Association |
| Source: |
The 26th European Conference on Power Electronics and Applications ; https://utc.hal.science/hal-05105944 ; The 26th European Conference on Power Electronics and Applications, GDR SEEDS France & EPE Association, Mar 2025, Paris, France. ⟨10.34746/epe2025-0144⟩ ; https://epe2025.com/ |
| Publisher Information: |
CCSD |
| Publication Year: |
2025 |
| Collection: |
Université de Technologie de Compiègne: HAL |
| Subject Terms: |
Condition monitoring; AC-DC converter; Vibration; Multi-objective optimization; Physics-Informed Machine Learning; [SPI]Engineering Sciences [physics]; [SPI.NRJ]Engineering Sciences [physics]/Electric power |
| Subject Geographic: |
Paris; France |
| Description: |
International audience ; Condition monitoring of power converters is vital but challenging due to sensor invasiveness and high computational demands. This paper introduces a non-invasive ultrasound-based approach using CNN autoencoders with a physics informed loss function for efficient data compression, outperforming traditional wavelet methods, enhancing data storage needs, and improving diagnostic feature extraction. |
| Document Type: |
conference object |
| Language: |
English |
| DOI: |
10.34746/epe2025-0144 |
| Availability: |
https://utc.hal.science/hal-05105944; https://utc.hal.science/hal-05105944v1/document; https://utc.hal.science/hal-05105944v1/file/0144-epe2025-full-17585544.pdf; https://doi.org/10.34746/epe2025-0144 |
| Rights: |
http://creativecommons.org/licenses/by/ ; info:eu-repo/semantics/OpenAccess |
| Accession Number: |
edsbas.A73078BC |
| Database: |
BASE |