Abstract
Artificial Bee Colony (ABC) is a population-based algorithm that employed the natural metaphors, based on foraging behavior of honey bee swarm. In ABC algorithm, there are three categories of bees. Employed bees select a random solution and apply a random neighborhood structure (exploration process), onlooker bees choose a food source depending on a selection strategy (exploitation process), and scout bees involves to search for new food sources (scouting process). In this paper. firstly we introduce a disruptive selection strategy for onlooker bees, to improve the diversity of the population and the premature convergence, and also a local search (i.e. simulated annealing) is introduced, in order to attain a balance between exploration and exploitation processes. Furthermore, a self adaptive strategy for selecting neighborhood structures is added to further enhance the local intensification capability. Experimental results show that the hybrid ABC with disruptive selection strategy outperforms the ABC algorithm alone when tested on examination timetabling problems.