TY - GEN
T1 - Development of Encoder-Decoder Predicting Search Process of Level-set Method in Magnetic Circuit Design
AU - Kawamata, Ryota
AU - Wakao, Shinji
AU - Murata, Noboru
AU - Okamoto, Yoshifumi
N1 - Funding Information:
ACKNOWLEDGEMENT A part of this work was supported by JSPS KAKENHI Grant Number 19H02132.
Publisher Copyright:
© 2019 IEEE.
PY - 2019/7
Y1 - 2019/7
N2 - The finite element analysis (FEA) of magnetic field generally requires a lot of calculation time. Especially, design optimization methods such as the level-set method with FEA result in large computational effort to find better solution. In this paper, we propose a novel method of precisely and quickly reproducing the conventional optimization steps by means of Convolutional Neural Network (CNN) and Long Short-term Memory (LSTM). The developed method enables us to implement high speed search of solution, which means the possibility of effective optimization with various initial conditions for better solution. Finally, we evaluate calculation time and computational accuracy of the proposed method by using a magnetic circuit design model.
AB - The finite element analysis (FEA) of magnetic field generally requires a lot of calculation time. Especially, design optimization methods such as the level-set method with FEA result in large computational effort to find better solution. In this paper, we propose a novel method of precisely and quickly reproducing the conventional optimization steps by means of Convolutional Neural Network (CNN) and Long Short-term Memory (LSTM). The developed method enables us to implement high speed search of solution, which means the possibility of effective optimization with various initial conditions for better solution. Finally, we evaluate calculation time and computational accuracy of the proposed method by using a magnetic circuit design model.
KW - Design optimization
KW - convolutional neural network
KW - level set method
KW - long short-term memory
KW - magnetic circuit
UR - http://www.scopus.com/inward/record.url?scp=85077946598&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85077946598&partnerID=8YFLogxK
U2 - 10.1109/COMPUMAG45669.2019.9032835
DO - 10.1109/COMPUMAG45669.2019.9032835
M3 - Conference contribution
AN - SCOPUS:85077946598
T3 - COMPUMAG 2019 - 22nd International Conference on the Computation of Electromagnetic Fields
BT - COMPUMAG 2019 - 22nd International Conference on the Computation of Electromagnetic Fields
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 22nd International Conference on the Computation of Electromagnetic Fields, COMPUMAG 2019
Y2 - 15 July 2019 through 19 July 2019
ER -