Antenna optimization based on Artificial Neural Network

In literature heuristic algorithms have been successfully applied to a number of electromagnetic problems. There are a number of approaches to build up the associated cost functions, the most common one is to link with full-wave analysis. However, this modelling method always leads to the point of complexity and high computational expense. Artificial Neural Network is one of the most robust biological inspired technique. In this article, a fast and accurate model is trained to replace the full-wave analysis in optimizing the bandwidth of a microstrip antenna. The comparison between ANN substitution model and full-wave characterization shows significant improvements in time convergence and computational cost. To verify the capability of proposed model, all these concepts are included in a case study of a rectangular ring antenna with proximity-coupled feed antenna.