Fast and efficient calculation of the multilayered shielded green's functions employing neural networks

Abstract

In this paper, neural networks are used to efficiently calculate the multilayered media boxed Green’s functions needed in integral equation (IE) formulations. The analysis of complex multilayered shielded circuits is computationally time consuming, due to the need to calculate the multilayered-media boxed Green’s functions. Using neural networks as radial basis function networks, the boxed Green’s functions can be calculated quickly, thus greatly reducing the computational time associated with the analysis of practical circuits. Once the neural network is trained with a known set of pairs of inputs and outputs, new outputs are quickly calculated, thus increasing the efficiency of the IE method.