Artificial Bee Colony (ABC) is an effective swarm optimization method featured with higher global search ability, less control parameters and easier implementation compared to other population-based optimization methods. Although ABC works well at exploration, its main drawback is poor exploitation affecting the convergence speed in some cases. In this paper, an efficient ABC-based optimization method is proposed to deal with high dimensional optimization tasks. The proposed method performs two modifications to the original ABC in order to improve its performance. First, it employs a chaos system to generate initial individuals, which are fully diversified in the search space. A chaos-based search method is used to find new solutions during ABC search process to enhance the exploitation capability of the algorithm and avoid premature convergence. Second, it incorporates a new search mechanism to improve the exploration ability of ABC. Experimental results performed on benchmark functions reveals superiority of the proposed method over state-of-the-art methods.
