The Back Propagation Neural Network Model of Non-Periodic Defected Ground Structure

Abstract

Presently, electromagnetic field numerical value analysis methods such as finite difference time-domain (FDTD) method are generally used to calculate the DGS, although these methods are accurate, they are also computationally expensive. In this paper, a neural network model of a novel defected ground structure is established. Since the neural network model has the advantages of great precision and effectiveness, the developed design model can be used to take the place of the FDTD method of the DGS, being a kind of aid tool of circuit design. The neural network models of two different non-periodic DGS have been developed, at the same time the circuit of the according DGS is designed and manufactured. The result of computer simulation and product measurements are obtained to demonstrate the effectiveness of the method.