Artificial bee colony algorithm improved by centroid strategy

KemingChen, ChunpingWang

COMPUTER MODELLING & NEW TECHNOLOGIES 2014 18(12A) 71-78

Mathematics and Computer Science Institute, XinYu University, JiangXi, China

Artificial Bee Colony algorithm was the best optimal algorithm proposed by Karaboga in Karaboga, having the following advantages, such as good stability, excellent ability of solution, less control parameters, simple and easy achievable computation, etc. It also had some disadvantages, such as premature convergence in its later period, poor accuracy in its development, etc. Therefore, a new-typed centroid improvement strategy is introduced in the paper for improving the searching ability of the artificial bee colony algorithm. This research is experimented by the common six kinds of testing functions. The centroid strategy introduced in the paper can effectively enhance the searching ability of the artificial bee colony algorithm in terms of the results so that the continuous development of the algorithm in its later period cannot be prematurely converged and the majority of the testing functions can be obviously improved