On the Design of Multilayer Circular Microstrip Antenna using Artificial Neural Networks

In this paper, we propose an artificial neural network (ANN) as a design technique for multilayer circular microstrip antennas based on Levenberg Marquardt training algorithm for modelling, simulation and optimization. Levenberg – Marquart (LM) algorithm has been used to train the MultiLayer Perceptron Neural Networks (MLPNNs). In the design procedure, the feed forward network is defined as a synthesis ANN model. Analysis ANN model is used as the reverse side of the problem to calculate the antenna dimension. The results of neural models has been compared with the measurements and calculated results. The results calculated by ANN model are found very close to the reference results. The average % accuracy in resonant frequency of ANN model for circular microstrip antenna with and without cover, Spaced dielectric Antenna and microstrip antenna with two superstrates is 0.35 %, 0.065 %, 0.43 % and 0.066 % respectively. The proposed ANN model requires no complicated mathematical formulas and suitable for CAD applications to design wide band, and high gain antenna.