A new artificial bee colony algorithm for solving large-scale optimization problems

Abstract

Artificial bee colony (ABC) is an efficient global optimizer, which has bee successfully used to solve various optimization problems. However, most of these problems are low dimensional. In this paper, we propose a new multi-population ABC (MPABC) algorithm to challenge large-scale global optimization problems. In MPABC, the population is divided into three subpopulations, and each subpopulation uses different search strategies. During the search, all subpopulations exchange there best search experiences to help accelerate the search. Experimental study is conducted on ten global optimization functions with dimensions 50, 100, and 200. Results show that MPABC is better than three other ABC variants on all dimensions.