Artificial Bee Colony (ABC) Algorithm with Crossover and Mutation

Artificial bee colony (ABC) is a relatively new swarm intelligence based metaheuristic. It was successfully applied to various, mostly continuous, optimization problems. For all such heuristically guided search algorithms balance between exploitation and exploration is the determining factor for success. It is generally considered that in the ABC algorithm exploitation is performed by employed bees and onlookers, while scout bees do the exploration. We have shown previously that the situation is more complex. In this paper we describe modifications to the ABC algorithm based on genetic algorithm (GA) crossover and mutation operators. Such modifications applied to the creation of new candidate solutions improved performance of the algorithm, tested on standard constrained optimization benchmark functions. This opens a new approach to exploitation/exploration balance for the ABC algorithm. Key-Words: Artificial bee colony (ABC), Constrained optimization, Swarm intelligence, Nature inspired metaheuristics, Genetic algorithm operators