Optimization of an Antipodal Vivaldi Antenna using Synthesis Neural Networks and a Novel Genetic Algorithms Approach

Abstract

Two global optimization techniques, applicable for the optimization of several antennas and microwave components, are presented. The two techniques are explained together with a sample application, which is the optimization of an antipodal Vivaldi antenna. The first technique was found useful in optimizing the design aspects whose physical behavior is known, whereas the second was better for optimizing the design aspects for which the response behavior is unknown, but an adequate performance metric can be defined. The two methods are compared; the limitations and advantages of each method arc presented together with some other optimization techniques.