Özet:

Optimization algorithms are algorithms aimed at finding the best solution by eliminating the global optimum. The Bonobo Optimization (BO) algorithm is a multi-intelligence algorithm based on the mathematical modeling of bonobo social behaviors and reproductive strategies. Bonobos form groups of different sizes and characteristics within the community they live, separating each other for different purposes and re-united after a while. When studying the reproductive strategy of bonobos, it appears that they have adopted four different strategies, such as random coupling, restrictive coupling, consortium and non-group coupling. These natural behaviors of Bonobos have been used to solve various optimization problems. The most important feature that distinguishes BO from many other intelligence-based algorithms is the search engines’ updating mechanisms and related parameters and the method of choosing the partners. In this study BO was studied and the parameters used in the algorithm were re-produced using the chaotic maps of Chebyshev, Circle, Gauss, Iterative, Logistic and Tent. Eight quality test functions have been used to compare performance. According to this, the results obtained from new algorithms created using chaotic maps have been shown to be more efficient than the classic BO.

Anahtar Kelimeler:

Özet:

Optimization algorithms are algorithms that aim to find the approximate best solution by sacrificing the global optimum. The Bonobo Optimization Algorithm examined in this study is an algorithm based on herd intelligence and is based on mathematical modeling of bonobos' social behavior and reproductive strategies. Bonobos form groups of various sizes and characteristics within the community they live in, leaving each other for different purposes and reuniting after a while. When the reproductive strategy of bonobos is examined, it is seen that they adopt four different strategies such as random mating, restrictive mating, consortium and out-group mating. These natural behaviors of bonobos have been used to solve various optimization problems. The most important feature that distinguishes the Bonobo Optimization Algorithm from other algorithms based on swarm intelligence is the update mechanisms of the search agents and the parameters associated with them, and the selection method of mating partners. In this study, the Bonobo Algorithm was examined and the results obtained by reproducing the random parameters used in the algorithm with chaotic maps were evaluated. It was analyzed in this study and reconstructed with the Chebyshev, Circle, Gauss, Iterative, Logistics and Tent chaotic maps used here. Eight quality tests were used for their performance. Accordingly, chaotic maps seem to be more efficient than classical BO from the results obtained from the new models.