Transfer Fonksiyonları Kullanarak İkili Güve-Alev Optimizasyonu Algoritmalarının Geliştirilmesi ve Performanslarının Karşılaştırılması

çözümü da Bu ikili optimizasyon problemlerinin çözümü için metasezgisel algoritmaların kullanımı Literatürde yer alan metasezgisel algoritmaların çoğu, sürekli problemlerin çözümüne uygun bir yapıya sahip olduğu için bu algoritmaların ikili problemleri çözebilecek şekilde düzenlenmesi gerekir. Transfer fonksiyonları olarak isimlendirilen Many real-world problems such as power systems problems, network optimization, backpack problems are referred to as binary optimization problems. The solution of binary optimization problems with classical mathematical techniques often takes a long time or is not possible. For this reason, the use of metaheuristic algorithms for the solution of binary optimization problems is quite common. Since most of the metaheuristic algorithms in the literature have a structure suitable for solving continuous problems, these algorithms should be arranged in a way that can solve binary problems. It is possible to convert continuous algorithms to binary algorithms by means of some functions called transfer functions. In this study, 8 different algorithms were developed by arranging the Moth-Flame Optimization (GAO) algorithm, a nature-inspired metaheuristic algorithm proposed in recent years, with 8 different transfer functions. The developed algorithms were run on 15 different uncapacitated facility location problems taken from the OR-Library and evaluated according to an error metric called gap. When the results are examined, it is observed that the binary GAO algorithm developed with the S3 transfer function achieves the minimum gap value in 13 of 15 problems and is more successful than the algorithms developed with other transfer functions.