Comparison of various optimization strategies for Yagi-Uda antenna design
This paper examines several optimization techniques as they are applied to Yagi-Uda antenna design. Most of the techniques discussed are various forms of genetic algorithms (GAs). A gradient method is also used both alone and in conjunction with GAs. Statistics are gathered as each method is run multiple times for the problem of a five element array optimized for gain, VSWR, and front-to-back ratio over a 5% bandwidth. After comparing these statistics, the GA+Gradient method proved to be the most efficient of the methods examined. It also produced the best mean antenna fitness over 50 optimization runs.