Title: Enhancing the firefly algorithm through a cooperative coevolutionary approach: an empirical study on benchmark optimisation problems

Authors: Giuseppe A. Trunfio

Addresses: Department of Architecture, Planning and Design, University of Sassari, P.zza Duomo, 6, 07041 Alghero (SS), Italy

Abstract: In recent years, the firefly algorithm (FA) has been applied with success to many classes of optimisation problems. However, as is the case for all metaheuristic optimisation algorithms, also with FA can be observed a rapid deterioration of efficiency as the dimensionality of the search space increases. In this paper, we use a cooperative coevolutionary approach for enhancing FA with the aim of making it much more efficient in the case of search spaces with many dimensions. We assess the performance of the cooperative coevolutionary firefly algorithm (CCFA) through a computational study based on some significant benchmark functions with up to 1,000 dimensions. Moreover, we compare the proposed CCFA with two state-of-the-art algorithms for high-dimensional optimisation problems. According to our results, CCFA can lead to significantly improved solutions in comparison to the standard FA. In addition, we show that the CCFA computation time is significantly lower than that of FA.

Keywords: cooperative coevolution; large-scale optimisation; swarm intelligence; firefly algorithm; bio-inspired computation; multidimensional search spaces.

DOI: 10.1504/IJBIC.2014.060621

International Journal of Bio-Inspired Computation, 2014 Vol.6 No.2, pp.108 - 125

Received: 02 Oct 2013
Accepted: 31 Jan 2014
Published online: 27 Sep 2014 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article