Design Optimization of Ultra Wide Band Vivaldi Antenna Using Artificial Intelligence

MAHOUTİ P., KIZILAY A. , Tari O., Belen A., Belen M. A.

International Symposium of the Applied-Computational-Electromagnetics-Society (ACES), ELECTR NETWORK, 1 - 05 August 2021 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.1109/aces53325.2021.00164
  • Keywords: Frequency selective surface, Optimization, Surrogate modelling, Ultra wide band, Vivaldi antenna, MAXIMUM POWER DELIVERY, TAPERED SLOT ANTENNA, YIELD ESTIMATION, HIGH-GAIN, MICROWAVE, TRANSISTOR, REFLECTOR


In this work, computationally efficient design optimization of Frequency Selective Surface (FSS) loaded Ultra-wide band Vivaldi antenna is studied. The proposed design consists of a multi-layer FSS structure aimed for performance improvement of the Vivaldi design, which makes the design a multi-objective multi dimension optimization problem. For having a fast and accurate optimization process, a data driven surrogate model alongside of meta-heurists optimizer Honey Bee Mating Optimization had been used. The obtained results are compared with 3D EM full wave simulator. Results show, the obtained FSS loaded structure have enhanced directivity compared with the design without FSS structure, without any performance losses in the return loss characteristics. The FSS loaded Vivaldi antenna operates at 2-12 GHZ band with a maximum gain of 10 dBi at 10 GHz which makes the design a good solution for RADAR applications.