Adaptive Artificial Bee Colony for Numerical Optimization

Abstract

Artificial bee colony (ABC) is a population-based optimizer. It simulates bees' social behavior for searching better solutions in solution space. Either too large or too small colony size will influence ABC's solution searching performance directly. In order to deal with the problem, in this paper, an adaptive colony is proposed. The adaptive colony will join potential bees or eliminate redundant bees, according solution searching situation. In experiments, 10 test functions of CEC 2015 are adopted for testing proposed method and compare it with three ABC variants. From the results, it can be observed that the proposed method performs better than other three related works.