Title: Dynamic differential evolution with oppositional orthogonal crossover for large scale optimisation problems
Authors: Xiaogang Dong; Changshou Deng; Yucheng Tan
Addresses: School of Information Science and Technology, JiuJiang University, JiuJiang, China ' School of Information Science and Technology, JiuJiang University, JiuJiang, China ' School of Information Science and Technology, JiuJiang University, JiuJiang, China
Abstract: Differential evolution is a population-based optimisation algorithm and has been successfully applied in many fields. However, when tackling large-scale optimisation problem, it still encounters serious challenges. To meet these challenges, a dynamic differential evolution with oppositional orthogonal crossover is proposed in this paper. A new oppositional learning is proposed and then it is used in the orthogonal crossover to improve the exploitation ability of the dynamic differential evolution. During the evolution process in dynamic differential evolution, only one individual is randomly chosen to undergo this oppositional orthogonal crossover operation. Thirteen benchmark problems with 1,000 dimensions were used to evaluate its performance. The results show that the proposed method is very competitive in terms of solution quality obtained.
Keywords: large-scale optimisation; dynamic differential evolution; DDE; oppositional orthogonal crossover; OOC; new oppositional learning.
DOI: 10.1504/IJCSM.2017.088016
International Journal of Computing Science and Mathematics, 2017 Vol.8 No.5, pp.414 - 424
Received: 07 Jun 2016
Accepted: 31 May 2017
Published online: 14 Nov 2017 *