BÜYÜK ÖLÇEKLİ SÜREKLİ OPTİMİZASYON PROBLEMLERİ İÇİN ELİT BİREY TABANLI YAPAY ARI KOLONİSİ ALGORİTMASI

Abstract

As the size of the optimization problems grows, it becomes more difficult to solve. Swarm intelligence algorithms can be used to tackle these problems. One of the swarm intelligence algorithms is the Artificial Bee Colony (ABC) algorithm. In order to benefit from the artificial bee colony algorithm in large-scale optimization problems, some improvements are required in the original ABC algorithm. In this study, the improvements made for the ABC algorithm are defined in a new ABC algorithm that we call "Elite Individual Based Adaptive Artificial Bee Colony Algorithm". Unlike classical ABC algorithms, different search equations are used in employed and onlooker bees steps and elite individuals are used in these search equations. In addition, the algorithm performance is enhanced with a local search technique. Choosing the correct parameter values of algorithms has a great effect on the success of algorithms. In this study, using the irace tool, the parameters of the algorithm are tried to be adjusted in the best way. The algorithm we developed is tested on the SOCO11 benchmark function set, which includes large-scale continuous optimization functions. The results we obtained were compared with ABC algorithms and the algorithms that previously participated in SOCO11 competition, and successful results are obtained.

Keywords

• artificial bee colony
• swarm intelligence
• large-scale optimization
• continuous optimization
• soco11