Gurusamy Jeyakumar and Chinthamani Nathan S. Velayutham
Differential evolution, dynamic differential evolution, differentialmutation strategies, successful runs, probability of convergence
In this paper, we extend the dynamicity of differential evolu- tion (DE) proposed for DE/rand/1/bin and DE/best/1/bin to five more variants DE/rand/2, DE/best/2, DE/current-to-rand/1, DE/current-to-best/1 and DE/rand-to-best/1. We present an em- pirical, comparative performance, analysis of 14 variants of DE and dynamic differential evolution (DDE ) algorithms (7 variants with two crossovers – binomial and exponential) to solve unconstrained global optimization problems. The aim of this paper is to identify competitive DE and DDE variants which perform well on different problems, and to compare the performance of DDE variants with DE variants. The performance of 14 variants of DE and DDE are analyzed by implementing them on 14 test functions. The analysis (done based on mean objective function value, probability of con- vergence and success performance) shows the superiority of DDE variants and identifies the competitive DE and DDE variants.
Important Links:
Go Back
