Wide-Band Near-Field Prediction of Unknown EM Source Based on Artificial Neural Network
- 20 September 2020
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-3
Abstract
In this paper, we propose a wideband radiation field prediction method based on artificial neural networks. This method makes up for the shortcomings of multiple measurements of single frequency points of traditional methods. At the same time, using the learning ability of ANN, we can only predict the radiation field by scanning the near-field magnitude. In the experiment, we scanned the near-field magnitude information of the radiation source at 100 MHz, 300 MHz, and 500 MHz. After training with the neural network, we can predict the near-field radiation of 100 MHz - 500 MHz, which proves the effectiveness of this method. This is a new approach to the Electromagnetic interference (EMI) problem.Keywords
Funding Information
- National Science Foundation
This publication has 5 references indexed in Scilit:
- A Miniature Multi-Component Probe for Near-Field ScanningIEEE Transactions on Antennas and Propagation, 2019
- An Equivalent Dipole Model Hybrid With Artificial Neural Network for Electromagnetic Interference PredictionIEEE Transactions on Microwave Theory and Techniques, 2019
- Machine-Learning-Based PML for the FDTD MethodIEEE Antennas and Wireless Propagation Letters, 2018
- Multivalued Neural Network Inverse Modeling and Applications to Microwave FiltersIEEE Transactions on Microwave Theory and Techniques, 2018
- An Improved Dipole-Moment Model Based on Near-Field Scanning for Characterizing Near-Field Coupling and Far-Field Radiation From an ICIEEE Transactions on Electromagnetic Compatibility, 2012