Application of neural networks to analyses of nonlinearly loaded antenna arrays including mutual coupling effects

In this paper, radial basis functions based neural networks (RBF-NN) are applied to the scattering of finite and infinite nonlinearly loaded antenna arrays including mutual coupling effects. The nodes in the input layer represent the parameters of antenna arrays or magnitudes of incident fields. There exist some nodes in the hidden layer for nonlinear mapping. The nodes in the output layer represent the magnitude of voltage at the input terminals of antennas at different harmonic frequencies. Numerical examples show that the scattering responses predicted by the trained RBF-NN models are very consistent with those calculated from the harmonic balance techniques. The trained RBF-NN models for the scattering of nonlinearly loaded antenna arrays are very efficient and the array mutual coupling effects are included.