2026-07-14T19:56:23Zhttps://ebiltegia.mondragon.edu/oai/request

oai:ebiltegia.mondragon.edu:20.500.11984/146242026-07-13T07:46:27Zcom_20.500.11984_1143com_20.500.11984_14090col_20.500.11984_1148

Arruti Romero, Asier Agote San Sebastian, Anartz Aizpuru, Iosu Mazuela, Mikel 2026-07-13T07:46:26Z 2026-07-13T07:46:26Z 2026 1048-2334 https://katalogoa.mondragon.edu/janium-bin/janium_login_opac.pl?find&ficha_no=202315 https://hdl.handle.net/20.500.11984/14624 This work focuses on addressing limitations and improving MU’s MagNet Challenge 2023 core loss model. MU’s core loss model is presented and explained, which formalizes and standardizes the model. Then, key limitations detected after the results of the MagNet Challenge are discussed. Some implementation issues of the model are fixed, which improves the performance of the model. In addition, upgrades to the model based on physical insight and ties with previous works are made, further improving the performance of the model. At the end, the fixed and improved model is about 60 % better accuracy and 40 % lower number of parameters. The model is designed to ease its adoption, and steps to facilitate replicating this work are taken. eng © 2026 IEEE Core Losses Ferrite Magnetic Devices ODS 7 Energía asequible y no contaminante MagNet Challenge 2023 Revisited: Beyond the Bounds of Equation-Based Core Loss Modeling