A Hybrid Method to Design Wire Antennas: Design and optimization of antennas using artificial intelligence

Abstract

Antenna design involves a set of complex requirements that may conflict. Some of these requirements are frequently expressed in terms of antenna parameters, such as radiation pattern, voltage standing wave ratio (VSWR) or impedance. In general, antenna requirements are specified for a frequency band or a set of frequencies. In some cases, the designer may create an antenna draft by trial and error until some performance goals are reached. Occasionally, this process may be simple, but in others, the number of variables makes it difficult to easily meet all of the goals. Alternatively, antenna design can be seen as an intricate error surface in an N-dimensional space. In this space, moving the solution over its surface in one direction may be good for some of the goals, but bad for others. Regularly, this error surface is difficult to travel through for most optimization algorithms. In previous works, genetic algorithms (GAs) have been used to assist the designer in this complicated process. This article proposes a hybrid method using artificial intelligence to design and optimize antennas. First, simulated annealing (SA) is used to draft the antenna. Second, the draft is improved until most of the performance goals are achieved.