Authors: Noorazliza Sulaiman; Junita Mohamad-Saleh; Abdul Ghani Abro
Addresses: School of Electrical and Electronic Engineering, Universiti Sains Malaysia, 14300, Nibong Tebal, Penang, Malaysia ' School of Electrical and Electronic Engineering, Universiti Sains Malaysia, 14300, Nibong Tebal, Penang, Malaysia ' College of Engineering, King Saud University Al Muzahimiyah Campus, 12372, Riyadh, Saudi Arabia
Abstract: The simplicity and robustness of the artificial bee colony (ABC) algorithm has attracted the attention of optimisation researchers. Although ABC has fewer tuned parameters, making it an easy-to-use tool, it has shown better performance than other prominent optimisation algorithms such as differential evolution (DE), evolutionary algorithms (EA) and particle swarm optimisation (PSO) algorithms at solving optimisation problems. Despite these advantages, researchers have found that the standard ABC actually suffers from slow convergence speed on unimodal functions and is often trapped in local minima of multimodal functions. Most problematically, it does not balance the exploitation and exploration stages, leading to various inefficiencies in terms of capability. This paper presents a new ABC variant referred to as JA-ABC4b, which has been formulated to balance exploitation and exploration in order to boost optimisation performance. JA-ABC4b has been experimentally tested on 27 benchmark functions and economic environmental dispatch (EED) problems. The results have revealed a robust performance of JA-ABC4b in comparison to other existing ABC variants and other optimisation algorithms.
Keywords: artificial intelligence; artificial bee colony; ABC; swarm intelligence-based algorithm; optimisation algorithm; economic environmental dispatch; EED.
International Journal of Bio-Inspired Computation, 2017 Vol.10 No.2, pp.99 - 108
Accepted: 25 Nov 2015
Published online: 28 Jul 2017 *