Детайли за публикацията
(Publication details)

Autors: Mateev, V. M., Marinova, I. Y.
Title: Machine Learning in Magnetic Field Calculations
Keywords: Bench-mark problems, Dirichlet boundary, Distributed excitation, Finite element method models, Harmonic magnetic fieldMachine learning approaches, Magnetic field calculations, Non-homogeneous

Abstract: Here is presented a machine learning approach for 2D steady-state and harmonic magnetic field calculations based on Poisson and Helmholtz equations for Dirichlet boundary problems. The approach is implemented on multilayer convolutional neural network trained over the Compumag 1b TEAM. benchmark problem variations dataset. Implementation is suitable for non-homogeneous magnetic properties domains and distributed excitation sources. Results accuracy is estimated in comparison with Finite Element Method model of the same problem


  1. Turner, L. (1985) Electromagnetic Effects in Magnetics, November
  2. Wiatowski, T., Bolcskei, H. A, 2018, Mathematical Theory of Deep Convolutional Neural Networks for Feature Extraction, IEEE Transactions on Information Theory, Volume 64(3), pp. pp. 1845-1866
  3. Yang, H.-F., Lin, K., Chen, C.-S., 2018, Supervised Learning of Semantics-Preserving Hash via Deep Convolutional Neural Networks, IEEE Transactions on Pattern Analysis and Machine Intelligence, Volume 40(2), pp. pp. 437-451, art. no. 7849132
  4. Ansys Inc, 2018, Testing Electromagnetic Analysis Methods (T. E. A. M.) International Compumag Society, Problems List, Ansys Inc, <https://www.lstc.com/applications/em/test_cases/team>, Дата на последен преглед (Last accessed on): 17.11.2020
  5. (2015) Best Practices for Convolutional Neural Networks Applied to Visual Document Analysis-Microsoft Research research. microsoft. com
  6. Deng, L., Yu, D., 2013, Deep learning: Methods and applications, Foundations and Trends in Signal Processing, Volume 7(3-4), pp. pp. 197-387

19th International Symposium on Electromagnetic Fields in Mechatronics, Electrical and Electronic Engineering, ISEF 2019, vol. ISEF 2019, issue 19, pp. Article number 9096969, 2019, France, IEEE Inc, DOI 10.1109/ISEF45929.2019.9096969

Copyright IEEE

Вид: пленарен доклад в международен форум, публикация в реферирано издание, индексирана в Scopus