An elitism based multi-objective artificial bee colony algorithm

Abstract

In this paper, we suggest a new multi-objective artificial bee colony (ABC) algorithm by introducing an elitism strategy. The algorithm uses a fixed-size archive that is maintained based on crowding-distance to store non-dominated solutions found during the search process. In the proposed algorithm, an improved artificial bee colony algorithm with an elitism strategy is adopted for the purpose of avoiding premature convergence. Specifically, the elites in the archive are selected and used to generate new food sources in both employed and onlooker bee phases in each cycle. To keep diversity, a member located at the most crowded region will be removed when the archive overflows. The algorithm is very easy to be implemented and it employs only a few control parameters. The proposed algorithm is tested on a wide range of multi-objective problems, and compared with other state-of-the-art algorithms in terms of often-used quality indicators with the help of a nonparametric test. It is revealed by the test procedure that the algorithm produces better or comparable results when compared with other well-known algorithms, and it can be used as a promising alternative tool to solve multi-objective problems with the advantage of being simple and effective.