An Adaptive Evolutionary Neural Network-Based Optimization Design Method for Wideband Dual-Polarized Antennas
Jinxin Li, Wenlu Qiu, Pei Xiao, Zhu Liu, Gaosheng Li, William T. Joines, Shaolin Liao
Abstract
In order to improve the efficiency and accuracy of antenna optimization, an adaptive evolutionary neural network (AENN)-based optimization design method for the wideband dual-polarized antenna is proposed. The proposed AENN is composed of two networks. The first network is used to predict the structural parameters of the antenna, and the second network uses the structural parameters as input to verify the results of the first network. This method only needs a small number of datasets to pretrain the network, and then, through the advanced workflow, while updating the dataset, the two networks can evolve iteratively, and finally, the antenna structural parameters under the expected electromagnetic response can be obtained. Taking two wideband dual-polarized antennas as numerical examples, this method is compared with other advanced antenna optimization methods, and the effectiveness of this method is verified.