An improved artificial bee colony algorithm for constrained optimization

Abstract

Artificial Bee Colony algorithm (ABC) is one of the most popular swarm intelligence algorithms possessing few control parameters and being competitive with other population-based algorithms. However, there is still an insufficiency in this algorithm regarding its convergence behavior. This algorithm is good at exploration but poor at exploitation and yet tackling the issue becomes more challenging if the problem involves constraints. In this research, an improved constrained ABC (iABC) algorithm is proposed to address this class of optimization problems. The modifications that have been introduced in iABC include a novel chaotic approach to generate initial population and two new search equations to enhance exploitation ability of the algorithm. In addition, a new fitness mechanism, along with an improved probability selection scheme has been devised to exploit both feasible and informative infeasible solutions. The proposed algorithm has been tested using CEC2006 benchmark suites. The performance of the iABC algorithm has been compared against the state of the art constrained ABC algorithms. According to the experimental results the proposed algorithm demonstrates a comparative performance and in some cases superior to the algorithms under study.