Deep Learning for Magnetic Field Estimation

• Published in: IEEE transactions on magnetics Vol. 55; no. 6; pp. 1 - 4
• Main Authors: Khan, Arbaaz, Ghorbanian, Vahid, Lowther, David
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.06.2019; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Artificial neural networks; Coil; Coils; Computational modeling; Computer simulation; convolutional neural networks; Deep learning; deep learning (DL); Design optimization; Empirical equations; Estimation; Finite element method; finite-element analysis (FEA); Geometry; Ground truth; Kernel; magnetic field; Magnetic fields; Magnetism; Mapping; Model accuracy; motor; Permanent magnets; Predictive models; Probabilistic models; Topology; Training; transformer; Uncertainty
• ISSN: 0018-9464; 1941-0069
• DOI: 10.1109/TMAG.2019.2899304

This paper investigates the feasibility of novel data-driven deep learning (DL) models to predict the solution of Maxwell's equations for low-frequency electromagnetic (EM) devices. With ground truth (empirical evidence) data being generated from a finite-element analysis solver, a deep convolutional neural network is trained in a supervised manner to learn a mapping for magnetic field distribution for topologies of different complexities of geometry, material, and excitation, including a simple coil, a transformer, and a permanent magnet motor. Preliminary experiments show DL model predictions in close agreement with the ground truth. A probabilistic model is introduced to improve the accuracy and to quantify the uncertainty in the prediction, based on Monte Carlo dropout. This paper establishes a basis for a fast and generalizable data-driven model used in the analysis, design, and optimization of EM devices.