Neural Network Approach to the Solution of Constructive Synthesis Problems of Active Phased Antenna Arrays

Abstract

The neural network approach to the solution of problems of constructive synthesis of the transmitting active phased antenna arrays is proposed. The considered antennas are composed of amplifiers with one or more nominal values and narrow ranges of transfer coefficients. The structure of a neural network consisting of a classifying neural network and several approximating neural networks is substantiated. The algorithm of neural network training with preliminary setting of the classifying part is proposed. Examples of solving problems of constructive synthesis, with different indicators of the quality of neural network training are given.