Design of electrically loaded wire antennas using genetic algorithms

Abstract

A novel antenna design procedure based on genetic algorithm (GA) driven optimization is proposed and applied to the synthesis of wire antennas loaded with lumped components. Loading circuit parameters, locations of the loads along the antenna, as well as matching network parameters, are optimized simultaneously. A computational scheme based on the Shernian-Morrison-Woodbury formula for the fast evaluation of the antenna performances for many distinct loading configurations is developed. The GA iteratively guides a population of randomly selected design candidates toward the optimal solution. The success of the proposed procedure is demonstrated through its application to the design of efficient ultra-broadband antennas and their corresponding matching networks. ©1996 IEEE.