Deep-Learning Based Transient Identification in Switched-Mode Power Supplies Conducted Emissions
- Resource Type
- Conference
- Authors
- Simonazzi, Mattia; Sandrolini, Leonardo; Iotti, Marcello; Mariscotti, Andrea
- Source
- 2022 International Symposium on Electromagnetic Compatibility – EMC Europe Electromagnetic Compatibility – EMC Europe, 2022 International Symposium on. :410-414 Sep, 2022
- Subject
- Aerospace
Bioengineering
Communication, Networking and Broadcast Technologies
Components, Circuits, Devices and Systems
Computing and Processing
Fields, Waves and Electromagnetics
Power, Energy and Industry Applications
Transportation
Time-frequency analysis
Switched mode power supplies
Europe
Artificial neural networks
Power system harmonics
Electromagnetic compatibility
Harmonic analysis
Neural Networks
Artificial Intelligence
Deep Learning
Electromagnetic Compatibility
Conducted Emissions
- Language
- ISSN
- 2325-0364
Conducted emissions (CE) caused by Switched-Mode Power Supplies (SMPSs) present harmonic and interharmonic distortion that occur in a wide range of frequencies and usually reveal a nonstationary behaviour. This requires long and complicated measures to ensure all the transient components to be correctly assessed. The analysis and classification of SMPS CE is addressed by employing an artificial neural network (ANN), with the aim of discriminate the part of the measured disturbance that is strongly affected by transient components and highlight the most relevant features of the CE spectrum. Thus, the subsequent frequency analysis can be performed on a smaller data set, allowing savings in time and computational efforts.