An Improved Artificial Bee Colony Algorithm with Non-separable Operator

摘要

Artificial Bee Colony (ABC) algorithm, motivated by the foraging behavior of honey bee swarm, has been shown to be competitive with other conventional nature inspired optimization algorithms. However, it has been found that the search mechanism using one element perturbation operator limits the algorithm's search ability in some cases. Therefore, we propose an improved ABC algorithm by embedding a non-separable operator and the gbest-guided operator in employed bee phase and onlooker bee phase, respectively, to balance the search performance on separable problem and non-separable problem. The effectiveness of the proposed ABC is analyzed on a standard benchmark suite consisting of eight functions. The undertaken study shows that the proposed ABC scheme exhibits a better performance compared to canonical ABC and its variant and is competitive with classic Differential Evolution (DE).