A Novel Artificial Bee Colony Algorithm with an Overall-Degradation Strategy and Its Performance on the Benchmark Functions of CEC 2014 Special Session
Bai Li · Automation Control and Intelligent Systems · 2014
The artificial bee colony (ABC) algorithm has been a well-known swarm intelligence algorithm, which assimilates the cooperating behavior of bees when seeking for nectar sources. Aiming to improve the conventional ABC algorithm, we focus on the re-initialization phase. In this paper, an overall-degradation-oriented artificial bee colony (OD-ABC) algorithm is proposed, pursuing to fight against premature convergence. This is achieved through re-initializing majority of the employed bees at one time, rather than generating at most one scout bee in each iteration. In this work, our OD-ABC algorithm is compared against the conventional ABC algorithms using 24 benchmark functions that origin from the CEC 2014’s competition on single objective real-parameter numerical optimization. The numerical results show that the OD-ABC algorithm is effective and thus can be employed to fight against premature convergence.