A Sampling Method for Neural Network-Based Antenna Modeling and Optimization

In this paper, a novel sampling method for antenna modeling and optimization is proposed. In this proposed method, the global samples are obtained by the global sampling method and the local samples are obtained by the local sampling method. The global samples are chosen based on the sample distribution to ensure the uniformity of the samples in the design space. The local samples are obtained by using particle swarm optimization to enhance the accuracy of the surrogate model in specific regions. The surrogate model based on artificial neural networks is utilized to obtain the optimal design parameters. The validity of the proposed sampling method is demonstrated through experimentation on a cavity-backed slot antenna. The results show that the proposed sampling method obtains the ANN model with good performance and can achieve optimization of the antenna.