Error prediction in electromagnetic simulations using machine learning


KARAOSMANOGLU B., ERGÜL Ö. S.

2019 IEEE International Symposium on Antennas and Propagation and USNC-URSI Radio Science Meeting, APSURSI 2019, Georgia, United States Of America, 7 - 12 July 2019, pp.787-788 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Volume:
  • Doi Number: 10.1109/apusncursinrsm.2019.8889056
  • City: Georgia
  • Country: United States Of America
  • Page Numbers: pp.787-788

Abstract

© 2019 IEEE.We present a novel approach of using deep convolutional neural networks (CNN) to predict electromagnetic scattering errors in iterative solutions of electrically large three-dimensional objects. Deep CNN models are constructed and trained by using surface current images to predict far-zone scattering errors. Numerical experiments demonstrate successful predictions with more than 95% accuracy. The constructed models can be useful to quickly assess the accuracy of candidate solutions of current distributions via their images.