An improved Artificial Bee Colony algorithm with incorporating information of qualified solutions

Artificial Bee Colony (ABC) is a metaheuristic algorithm with proper ability in solving optimization problems. However, its performance can be improved by setting a better balance between exploitation and exploration. In this study, by changing the search pattern of neighborhood and incorporating the information of a set of qualified solutions into the creating process of candidate solutions, the balance between exploitation and exploration would improve. This change is in a way that in addition to improvement of exploration, the capability of employed and onlooker bees to search around proper solutions is utilized properly. Experiments are conducted on 22 different benchmark functions including standard, shifted, rotated, and shifted-rotated multimodal and unimodal problems. The results confirm superiority of the proposed algorithm compared to standard ABC and some new versions of it.