A neural network combined with a three-dimensional finite element method applied to optimize eddy current and temperature distributions of traveling wave induction heating system

Youhua Wang, Junhua Wang, Siu Lau Ho, Lingling Pang, Weinong Fu

Research output: Journal article publicationJournal articleAcademic researchpeer-review

3 Citations (Scopus)

Abstract

In this paper, neural networks with a finite element method (FEM) were introduced to predict eddy current distributions on the continuously moving thin conducting strips in traveling wave induction heating (TWIH) equipments. A method that combines a neural network with a finite element method (FEM) is proposed to optimize eddy current distributions of TWIH heater. The trained network used for tested examples shows quite good accuracy of the prediction. The results have then been used with reference to a double-side TWIH in order to analyze the distributions of the magnetic field and eddy current intensity, which accelerates the iterative solution process for the nonlinear coupled electromagnetic matters. The FEM computation of temperature converged conspicuously faster using the prediction results as initial values than using the zero values, and the number of iterations is reduced dramatically. Simulation results demonstrate the effectiveness and characteristics of the proposed method.
Original languageEnglish
Article number07E522
JournalJournal of Applied Physics
Volume109
Issue number7
DOIs
Publication statusPublished - 1 Apr 2011

ASJC Scopus subject areas

  • Physics and Astronomy(all)

Cite this