Abstract:
Artificial Bee Colony (ABC) algorithm inspired by the foraging behaviors of
real honey bees is one of the most important swarm intelligence based optimization
algorithms. When considering the robust and phase divided structure of the ABC
algorithm, it is clearly seen that ABC algorithm is intrinsically suitable for
parallelization. In this paper, we proposed a new emigrant creation strategy for parallel
ABC algorithm. The proposed model named order based emigrant creation strategy
depends on sending best food source in a subcolony after modifying it with another
food source chosen sequentially from the same subcolony at each migration time.
Experimental studies on a set of numerical benchmark functions showed that parallel
ABC algorithm powered by the newly proposed model significantly improved quality of
the final solutions and convergence performance when compared with standard serial
ABC algorithm and parallel ABC algorithm for which the best food sources in the
subcolonies directly used as emigrants.
