Intermittent fault diagnosis based on EMD decomposition and convolutional neural network
The paper introduces an intermittent fault diagnosis method for analog circuits using EMD and CNN. The proposed method obtains the circuit output characteristics under different intermittent fault type using SPICE simulation. Then, EMD is used to extract the features and finally, the CNN model for fault detection and localization is built. The effectiveness of the method is validated through experiments in a three-phase SVPWM circuit.
Authors
- Shenhang Wang
- Ruishi Lin (ORCID: https://orcid.org/0009-0003-5361-7135)
- Gang Xiang (ORCID: https://orcid.org/0000-0002-6278-3085)
- Yu Peng (ORCID: https://orcid.org/0000-0003-3244-838X)
- Yuhang Liu (ORCID: https://orcid.org/0000-0001-6500-3278)
- Zhiming Yang (ORCID: https://orcid.org/0000-0001-8614-722X)
Institutions
- Harbin Institute of Technology (CN)
- Beijing Aerospace Flight Control Center (CN)
Publication Details
- Published
- 2024-12-29
- DOI
- https://doi.org/10.1109/iceace63551.2024.10898531
- Primary Topic
- Advanced Computational Techniques and Applications
- Type
- article
- Field-Weighted Citation Impact
- 0.00