An Adaptive Artifical Bee Colony Algorithms Based on Global Best Guide

The artificial bee colony (ABC) algorithm is one of the popular swarm intelligence algorithms that is used to solve many mathematical problems and real-world optimization problems. However, ABC sometimes may be slow to converge and good at exploration but poor at exploitation. In order to improve the algorithm performance, we propose a modified algorithm called adaptive and global artificial bee colony (AGABC) algorithm based on global best guide. In this paper, we propose the new method for updating solution of the onlooker bees. In our method, the best solution at current iteration is used to enhance the ability of exploitation. In addition, the initial population is generated by using a chaotic system to enhance the global convergence. Moreover, AGABC algorithm pays attention to global exploration at early phase and then turn to exploitation coming to end. We experiment the performance of our proposed method on two sets of problems: numerical benchmark functions and the chaotic time series prediction via Volterra method application. Experimental results show that the proposed method is able to attain higher quality solutions with faster convergence than the original ABC or some improved ABC algorithms.