A Neural Network Model for Effective Dielectric Constant Prediction of a Two Layered Microstrip Transmission Line

In this paper, a neural network model is developed to calculate the effective dielectric constant of a two-layered microstrip transmission line. The neural network code is developed in Keras. The model is trained using effective dielectric constant of a two-layered dielectric substrate and dielectric values of substrates. The data required for the training of the network is obtained from a commercial EM solver TaraNG. The suitable network structure is 2-10-10-1. The predicted results are well matched with commercial EM TaraNG solver results.