An Astute Artificial Bee Colony Algorithm

Abstract

Artificial bee colony (ABC) algorithm is one of the most popular optimization methods for global optimization over real-valued parameters. Though it has been shown very competitive to other natureinspired methods, it suffers from some challenging problems, e.g., slow convergence speed while solving unimodal problems, local optima stagnation (premature convergence) while dealing with the complex multimodal problems, and scalability problem in case of high dimensional problems. In order to circumvent these problems, we propose a new variant of the ABC, called Astute Artificial Bee Colony (AsABC) algorithm, which is able to maintain a better trade-off between two conflicting aspects, exploration and exploitation in the search space. In AsABC, we model a new search behavior of the onlooker bees to foster the solutions towards better region and to make the algorithm scalable. Performance of the AsABC is evaluated on a test suite of 12 benchmark functions of three different categories: unimodal, multimodal, and rotated multimodal. Comprehensive benchmarking and comparison of the AsABC with three other state-of-the-art variants of the ABC demonstrate its superior performance in terms of solution quality, scalability, robustness, and convergence speed.