A Fast Systematic Approach for Microstrip Antenna Design and Optimization using ANFIS and GA

Abstract

In this paper an approach for optimizing the physical parameters of microstrip antennas is presented. Although this method can be used for optimizing any kind of antennas, our focus is on ultra wide-band (UWB) microstrip antennas. Adaptive neuro-fuzzy inference systems (ANFIS) trained by hybrid learning algorithm are used for estimating the properties of the desired antennas. The data required for training these networks is extracted from HFSS. By introducing a proper objective function and applying genetic algorithm (GA) to it, the physical parameters of the antenna are determined for bandwidth maximization. In the optimization process, instead of running HFSS repetitively, trained ANFISs are used for antenna characteristics estimation, and therefore the optimization time in this method is much lower than other techniques. Comparison of the parameters obtained by our method and those reported previously shows the accuracy and efficiency of the method.