MagNet Challenge for Data-Driven Power Magnetics Modeling

Minjie Chen; Haoran Li; Shukai Wang; Thomas Guillod; Diego Serrano; Lizhong Zhang; Tom McKeague; Navid Rasekh; Binyu Cui; Jun Wang; Song Liu; Alfonso Martinez; Shuai Jiang; David Perreault; Dragan Maksimovic; Ron Hui; Johann Kolar; David Shumacher; Charles Sullivan

doi:10.1109/OJPEL.2024.3469916

MagNet Challenge for Data-Driven Power Magnetics Modeling

Minjie Chen^*, Haoran Li, Shukai Wang, Thomas Guillod, Diego Serrano, Lizhong Zhang, Tom McKeague, Navid Rasekh, Binyu Cui, Jun Wang, Song Liu, Alfonso Martinez, Shuai Jiang, David Perreault, Dragan Maksimovic, Ron Hui, Johann Kolar, David Shumacher, Charles Sullivan

^*Corresponding author for this work

Research output: Contribution to journal › Article (Academic Journal) › peer-review

4 Citations (Scopus)

Abstract

This paper summarizes the main results and contributions of the MagNet Challenge 2023, an open-source research initiative for data-driven modeling of power magnetic materials. The MagNet Challenge has (1) advanced the stateof-the-art in power magnetics modeling; (2) set up examples for fostering an open-source and transparent research community; (3) developed useful guidelines and practical rules for conducting data-driven research in power electronics; and (4) provided a fair performance benchmark leading to insights on the most promising future research directions. The competition yielded a collection of publicly disclosed software algorithms and tools designed to capture the distinct loss characteristics of power magnetic materials, which are mostly open-sourced. We have attempted to bridge power electronics domain knowledge with state-of-the-art advancements in artificial intelligence, machine learning, pattern recognition, and signal processing. The MagNet Challenge has greatly improved the accuracy and reduced the size of data-driven power magnetic material models. The models and tools created for various materials were meticulously documented and shared within the broader power electronics community.

Original language	English
Journal	IEEE Open Journal of Power Electronics
DOIs	https://doi.org/10.1109/OJPEL.2024.3469916
Publication status	Published - 30 Sept 2024

Bibliographical note

Publisher Copyright:
© 2024 Institute of Electrical and Electronics Engineers Inc.. All rights reserved.

Keywords

open source
data-driven methods
machine learning
artificial intelligence
power magnetics
power ferrites

Access to Document

10.1109/OJPEL.2024.3469916Licence: CC BY-NC-ND

Handle.net

Persistent link

EPSRC Impact Acceleration Account - Exploratory Award
Wang, J. (Recipient), 2022
Prize: Prizes, Medals, Awards and Grants
Outstanding Performance Award 3rd Place, MagNet Challenge 2023
Wang, J. (Recipient), McKeague, T. (Recipient), Zhang, L. (Recipient), Cui, B. (Recipient) & Liu, S. (Recipient), 28 Feb 2024
Prize: Prizes, Medals, Awards and Grants

HPC (High Performance Computing) and HTC (High Throughput Computing) Facilities
Alam, S. R. (Manager), Williams, D. A. G. (Manager), Eccleston, P. E. (Manager) & Greene, D. (Manager)
Facility/equipment: Facility

Cite this

Chen, M., Li, H., Wang, S., Guillod, T., Serrano, D., Zhang, L., McKeague, T., Rasekh, N., Cui, B., Wang, J., Liu, S., Martinez, A., Jiang, S., Perreault, D., Maksimovic, D., Hui, R., Kolar, J., Shumacher, D., & Sullivan, C. (2024). MagNet Challenge for Data-Driven Power Magnetics Modeling. IEEE Open Journal of Power Electronics. https://doi.org/10.1109/OJPEL.2024.3469916

@article{81e61b330ec04438930d61e600b17683,

title = "MagNet Challenge for Data-Driven Power Magnetics Modeling",

abstract = "This paper summarizes the main results and contributions of the MagNet Challenge 2023, an open-source research initiative for data-driven modeling of power magnetic materials. The MagNet Challenge has (1) advanced the stateof-the-art in power magnetics modeling; (2) set up examples for fostering an open-source and transparent research community; (3) developed useful guidelines and practical rules for conducting data-driven research in power electronics; and (4) provided a fair performance benchmark leading to insights on the most promising future research directions. The competition yielded a collection of publicly disclosed software algorithms and tools designed to capture the distinct loss characteristics of power magnetic materials, which are mostly open-sourced. We have attempted to bridge power electronics domain knowledge with state-of-the-art advancements in artificial intelligence, machine learning, pattern recognition, and signal processing. The MagNet Challenge has greatly improved the accuracy and reduced the size of data-driven power magnetic material models. The models and tools created for various materials were meticulously documented and shared within the broader power electronics community.",

keywords = "open source, data-driven methods, machine learning, artificial intelligence, power magnetics, power ferrites",

author = "Minjie Chen and Haoran Li and Shukai Wang and Thomas Guillod and Diego Serrano and Lizhong Zhang and Tom McKeague and Navid Rasekh and Binyu Cui and Jun Wang and Song Liu and Alfonso Martinez and Shuai Jiang and David Perreault and Dragan Maksimovic and Ron Hui and Johann Kolar and David Shumacher and Charles Sullivan",

year = "2024",

month = sep,

day = "30",

doi = "10.1109/OJPEL.2024.3469916",

language = "English",

journal = "IEEE Open Journal of Power Electronics",

issn = "2644-1314",

publisher = "Institute of Electrical and Electronics Engineers (IEEE)",

}

TY - JOUR

T1 - MagNet Challenge for Data-Driven Power Magnetics Modeling

AU - Chen, Minjie

AU - Li, Haoran

AU - Wang, Shukai

AU - Guillod, Thomas

AU - Serrano, Diego

AU - Zhang, Lizhong

AU - McKeague, Tom

AU - Rasekh, Navid

AU - Cui, Binyu

AU - Wang, Jun

AU - Liu, Song

AU - Martinez, Alfonso

AU - Jiang, Shuai

AU - Perreault, David

AU - Maksimovic, Dragan

AU - Hui, Ron

AU - Kolar, Johann

AU - Shumacher, David

AU - Sullivan, Charles

PY - 2024/9/30

Y1 - 2024/9/30

N2 - This paper summarizes the main results and contributions of the MagNet Challenge 2023, an open-source research initiative for data-driven modeling of power magnetic materials. The MagNet Challenge has (1) advanced the stateof-the-art in power magnetics modeling; (2) set up examples for fostering an open-source and transparent research community; (3) developed useful guidelines and practical rules for conducting data-driven research in power electronics; and (4) provided a fair performance benchmark leading to insights on the most promising future research directions. The competition yielded a collection of publicly disclosed software algorithms and tools designed to capture the distinct loss characteristics of power magnetic materials, which are mostly open-sourced. We have attempted to bridge power electronics domain knowledge with state-of-the-art advancements in artificial intelligence, machine learning, pattern recognition, and signal processing. The MagNet Challenge has greatly improved the accuracy and reduced the size of data-driven power magnetic material models. The models and tools created for various materials were meticulously documented and shared within the broader power electronics community.

AB - This paper summarizes the main results and contributions of the MagNet Challenge 2023, an open-source research initiative for data-driven modeling of power magnetic materials. The MagNet Challenge has (1) advanced the stateof-the-art in power magnetics modeling; (2) set up examples for fostering an open-source and transparent research community; (3) developed useful guidelines and practical rules for conducting data-driven research in power electronics; and (4) provided a fair performance benchmark leading to insights on the most promising future research directions. The competition yielded a collection of publicly disclosed software algorithms and tools designed to capture the distinct loss characteristics of power magnetic materials, which are mostly open-sourced. We have attempted to bridge power electronics domain knowledge with state-of-the-art advancements in artificial intelligence, machine learning, pattern recognition, and signal processing. The MagNet Challenge has greatly improved the accuracy and reduced the size of data-driven power magnetic material models. The models and tools created for various materials were meticulously documented and shared within the broader power electronics community.

KW - open source

KW - data-driven methods

KW - machine learning

KW - artificial intelligence

KW - power magnetics

KW - power ferrites

U2 - 10.1109/OJPEL.2024.3469916

DO - 10.1109/OJPEL.2024.3469916

M3 - Article (Academic Journal)

SN - 2644-1314

JO - IEEE Open Journal of Power Electronics

JF - IEEE Open Journal of Power Electronics

ER -

MagNet Challenge for Data-Driven Power Magnetics Modeling

Abstract

Bibliographical note

Keywords

Access to Document

Handle.net

Fingerprint

Prizes

EPSRC Impact Acceleration Account - Exploratory Award

Outstanding Performance Award 3rd Place, MagNet Challenge 2023

Equipment

HPC (High Performance Computing) and HTC (High Throughput Computing) Facilities

Cite this