Abstract: Artificial Bee Colony (ABC) algorithm is a new swarm intelligence technique inspired by the foraging behavior of a honeybee swarm.As the member of the artificial bee colony,scout bees carry out random search for discovering food sources.In order to investigate the effect of scout bees on the performance of ABC,the biological mechanism and main steps of ABC were analyzed,and then,different problem dimensions,population sizes,limit values and maximum cycle numbers were tested on the performance of ABC under the conditions of no scout bee,single scout bee and multi-scout bees conditions.Almost all the experimental results show that ABC with multi-scout bees outperforms ABC with single scout bee and ABC without scout bee on five well-known benchmark functions,meanwhile,ABC with single scout bee performs better than ABC without scout bee.Besides,the three algorithms have almost the same execution time under the same conditions due to the same order of their search complexity.These fully demonstrate that the random exploration process adopted by scout bees has positive effect on the performance of ABC.

Key words: artificial bee colony algorithm, scout bee, exploration process