Artificial Neural Network (ANN) Based Fast and Accurate Inductor Modeling and Design
This paper analyzes the potential of Artificial Neural Networks (ANNs) for the modeling and optimization of magnetic components and, specifically, inductors. After reviewing the basic properties of ANNs, several potential modeling and design workflows are presented. A hybrid method, which combines t...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2020-01-01
|
| Series: | IEEE Open Journal of Power Electronics |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9152082/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841533457312776192 |
|---|---|
| author | Thomas Guillod Panteleimon Papamanolis Johann W. Kolar |
| author_facet | Thomas Guillod Panteleimon Papamanolis Johann W. Kolar |
| author_sort | Thomas Guillod |
| collection | DOAJ |
| description | This paper analyzes the potential of Artificial Neural Networks (ANNs) for the modeling and optimization of magnetic components and, specifically, inductors. After reviewing the basic properties of ANNs, several potential modeling and design workflows are presented. A hybrid method, which combines the accuracy of 3D Finite Element Method (FEM) and the low computational cost of ANNs, is selected and implemented. All relevant effects are considered (3D magnetic and thermal field patterns, detailed core loss data, winding proximity losses, coupled loss-thermal model, etc.) and the implemented model is extremely versatile (<inline-formula><tex-math notation="LaTeX">$30$</tex-math></inline-formula> input and <inline-formula><tex-math notation="LaTeX">$40$</tex-math></inline-formula> output variables). The proposed ANN-based model can compute <inline-formula><tex-math notation="LaTeX">$50^{\prime}000$</tex-math></inline-formula> designs per second with less than <inline-formula><tex-math notation="LaTeX">$3 \%$</tex-math></inline-formula> deviation with respect to 3D FEM simulations. Finally, the inductor of a <inline-formula><tex-math notation="LaTeX">$2$</tex-math></inline-formula> kW DC-DC buck converter is optimized with the ANN-based workflow. From the Pareto fronts, a design is selected, measured, and successfully compared with the results obtained with the ANNs. The implementation (source code and data) of the proposed workflow is available under an open-source license. |
| format | Article |
| id | doaj-art-fc89c916951a4063bd85aea1fc727c43 |
| institution | Kabale University |
| issn | 2644-1314 |
| language | English |
| publishDate | 2020-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Open Journal of Power Electronics |
| spelling | doaj-art-fc89c916951a4063bd85aea1fc727c432025-01-16T00:02:24ZengIEEEIEEE Open Journal of Power Electronics2644-13142020-01-01128429910.1109/OJPEL.2020.30127779152082Artificial Neural Network (ANN) Based Fast and Accurate Inductor Modeling and DesignThomas Guillod0https://orcid.org/0000-0003-0738-5823Panteleimon Papamanolis1https://orcid.org/0000-0002-3732-2332Johann W. Kolar2https://orcid.org/0000-0002-6000-7402Power Electronic Systems Laboratory, ETH Zurich, Zurich, SwitzerlandPower Electronic Systems Laboratory, ETH Zurich, Zurich, SwitzerlandPower Electronic Systems Laboratory, ETH Zurich, Zurich, SwitzerlandThis paper analyzes the potential of Artificial Neural Networks (ANNs) for the modeling and optimization of magnetic components and, specifically, inductors. After reviewing the basic properties of ANNs, several potential modeling and design workflows are presented. A hybrid method, which combines the accuracy of 3D Finite Element Method (FEM) and the low computational cost of ANNs, is selected and implemented. All relevant effects are considered (3D magnetic and thermal field patterns, detailed core loss data, winding proximity losses, coupled loss-thermal model, etc.) and the implemented model is extremely versatile (<inline-formula><tex-math notation="LaTeX">$30$</tex-math></inline-formula> input and <inline-formula><tex-math notation="LaTeX">$40$</tex-math></inline-formula> output variables). The proposed ANN-based model can compute <inline-formula><tex-math notation="LaTeX">$50^{\prime}000$</tex-math></inline-formula> designs per second with less than <inline-formula><tex-math notation="LaTeX">$3 \%$</tex-math></inline-formula> deviation with respect to 3D FEM simulations. Finally, the inductor of a <inline-formula><tex-math notation="LaTeX">$2$</tex-math></inline-formula> kW DC-DC buck converter is optimized with the ANN-based workflow. From the Pareto fronts, a design is selected, measured, and successfully compared with the results obtained with the ANNs. The implementation (source code and data) of the proposed workflow is available under an open-source license.https://ieeexplore.ieee.org/document/9152082/Power convertersartificial neural networksfinite element analysisinductorsmachine learningmagnetic devices |
| spellingShingle | Thomas Guillod Panteleimon Papamanolis Johann W. Kolar Artificial Neural Network (ANN) Based Fast and Accurate Inductor Modeling and Design IEEE Open Journal of Power Electronics Power converters artificial neural networks finite element analysis inductors machine learning magnetic devices |
| title | Artificial Neural Network (ANN) Based Fast and Accurate Inductor Modeling and Design |
| title_full | Artificial Neural Network (ANN) Based Fast and Accurate Inductor Modeling and Design |
| title_fullStr | Artificial Neural Network (ANN) Based Fast and Accurate Inductor Modeling and Design |
| title_full_unstemmed | Artificial Neural Network (ANN) Based Fast and Accurate Inductor Modeling and Design |
| title_short | Artificial Neural Network (ANN) Based Fast and Accurate Inductor Modeling and Design |
| title_sort | artificial neural network ann based fast and accurate inductor modeling and design |
| topic | Power converters artificial neural networks finite element analysis inductors machine learning magnetic devices |
| url | https://ieeexplore.ieee.org/document/9152082/ |
| work_keys_str_mv | AT thomasguillod artificialneuralnetworkannbasedfastandaccurateinductormodelinganddesign AT panteleimonpapamanolis artificialneuralnetworkannbasedfastandaccurateinductormodelinganddesign AT johannwkolar artificialneuralnetworkannbasedfastandaccurateinductormodelinganddesign |