相关参考文献
注意:仅列出部分参考文献,下载原文获取全部文献信息。
Article
Engineering, Electrical & Electronic
Marco Baldan et al.
Summary: Physics-informed neural networks (PINNs) have been successfully used for solving direct problems in electromagnetism (EM). However, when dealing with inverse problems or parameter studies, PINNs usually require additional training for each new set of system parameters. To address this issue, we propose a hypernetwork (HNN) that takes system parameters as input and outputs the network weights of a PINN, enabling the fast solution of inverse problems. This approach is demonstrated in a coil optimal design task in magnetostatics.
IEEE TRANSACTIONS ON MAGNETICS
(2023)
Article
Computer Science, Interdisciplinary Applications
Piergiorgio Alotto et al.
Summary: The study aims to define a benchmark problem to test new regularization methods and solutions with different approaches.
COMPEL-THE INTERNATIONAL JOURNAL FOR COMPUTATION AND MATHEMATICS IN ELECTRICAL AND ELECTRONIC ENGINEERING
(2021)
Review
Chemistry, Multidisciplinary
Yanbin Li et al.
Summary: This paper reviews recent developments in design optimization methods for electromagnetic devices, focusing on machine learning techniques. It covers advancements in multi-objective, multidisciplinary, multilevel, topology, fuzzy, and robust design optimization, as well as the use of machine learning algorithms such as artificial neural networks, support vector machines, and deep learning. Future directions discussed include the application of cloud services and digital twin technology to improve manufacturing quality and reliability.
APPLIED SCIENCES-BASEL
(2021)
Article
Energy & Fuels
Elia Brescia et al.
Summary: A novel approach is proposed to minimize the cogging torque of permanent magnet machines with segmented stator cores, by designing multiple independent shapes of the tooth tips through topological optimization. A computationally-efficient heuristic approach based on genetic algorithms and artificial neural network-based surrogate models is utilized to find the optimal tooth tips shapes, which is validated through simulation studies. The proposed method proves to be superior compared to a conventional heuristic approach based on a genetic algorithm coupled to finite element analysis, and shows robustness in a sensitivity analysis on assembling and manufacturing tolerances.
Article
Engineering, Electrical & Electronic
Sami Barmada et al.
Summary: A topology optimization procedure is proposed and applied to the TEAM 25 problem involving a die press model with an electromagnet for orientation of magnetic powder. The relationship between the press shape and flux density in the cavity is simulated using finite element analysis and learned by a deep neural network model. The DNN is used as a surrogate model for optimization, leading to promising results in terms of improved accuracy and reduced time cost compared to the full FEA-based optimization approach.
IEEE TRANSACTIONS ON MAGNETICS
(2021)
Article
Engineering, Electrical & Electronic
Jaweria Amjad et al.
Summary: The paper explores utilizing knowledge of the forward operator to enhance generalization of deep learning approaches in inverse problems, providing generalization error bounds, a 'plug-and-play' regularizer, and efficient methods for bounding the Jacobians. The model-aware regularized deep learning algorithms demonstrated superiority in medical imaging tasks.
IEEE TRANSACTIONS ON SIGNAL PROCESSING
(2021)
Article
Engineering, Electrical & Electronic
Pan Zhang et al.
Summary: This paper discusses an unsupervised deep learning method for solving time domain electromagnetic simulations. By encoding initial conditions, boundary conditions, and Maxwell's equations as constraints during training, the method turns electromagnetic simulation into an optimization process. The high prediction accuracy of electromagnetic fields and computational efficiency in various media are demonstrated through numerical examples, validating the effectiveness of this approach for high-dimensional and multi-scale simulations.
IEEE JOURNAL ON MULTISCALE AND MULTIPHYSICS COMPUTATIONAL TECHNIQUES
(2021)
Article
Computer Science, Information Systems
Vivek Parekh et al.
Summary: The use of magneto-static finite element simulation in obtaining key performance indicators of electrical machine designs is time-consuming and computationally expensive. This study introduces a data-assisted deep learning-based meta-model to predict KPIs efficiently and accurately, reducing computational costs. The results demonstrate high prediction accuracy and the potential of image-based approach to minimize optimization time.
Editorial Material
Engineering, Electrical & Electronic
Dong Liang et al.
IEEE SIGNAL PROCESSING MAGAZINE
(2020)
Article
Engineering, Electrical & Electronic
Sami Barmada et al.
IEEE TRANSACTIONS ON MAGNETICS
(2020)
Article
Engineering, Electrical & Electronic
Jo Asanuma et al.
IEEE TRANSACTIONS ON MAGNETICS
(2020)
Review
Computer Science, Information Systems
Md Zahangir Alom et al.
Article
Engineering, Electrical & Electronic
Alessandro Formisano et al.
INTERNATIONAL JOURNAL OF APPLIED ELECTROMAGNETICS AND MECHANICS
(2019)
Article
Engineering, Electrical & Electronic
Arbaaz Khan et al.
IEEE TRANSACTIONS ON MAGNETICS
(2019)
Article
Engineering, Electrical & Electronic
Hidenori Sasaki et al.
IEEE TRANSACTIONS ON MAGNETICS
(2019)
Article
Engineering, Electrical & Electronic
Alice Lucas et al.
IEEE SIGNAL PROCESSING MAGAZINE
(2018)
Article
Engineering, Electrical & Electronic
Michael T. McCann et al.
IEEE SIGNAL PROCESSING MAGAZINE
(2017)
Article
Computer Science, Artificial Intelligence
Kyong Hwan Jin et al.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2017)
Article
Statistics & Probability
H Zou et al.
JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES B-STATISTICAL METHODOLOGY
(2005)