Novel differential evolution algorithm with spatial evolution rules

In order to reduce the pressure of parameter selection and avoid trapping into the local optimum,a novel differential evolution（ DE） algorithm without crossover rate is proposed. Through embedding cellular automata into the DE algorithm,those interactions among vectors are restricted within cellular...

Full description

Saved in:
Bibliographic Details
Published in高技术通讯（英文版） Vol. 23; no. 4; pp. 426 - 433
Main Author 丁青锋;Qiu Xiang
Format Journal Article
LanguageEnglish
Published School of Electrical and Automation Engineering, East China Jiaotong University, Nanchang 330013, P.R.China 01.12.2017
Subjects
selective pressure
differential evolution ( DE )
orthogonal crossover
balancing tradeoff
cellular automata
Online AccessGet full text
ISSN1006-6748
DOI10.3772/j.issn.1006-6748.2017.04.012

Cover

More Information
Summary:In order to reduce the pressure of parameter selection and avoid trapping into the local optimum,a novel differential evolution（ DE） algorithm without crossover rate is proposed. Through embedding cellular automata into the DE algorithm,those interactions among vectors are restricted within cellular structure of neighbors while the cell own evolution,which may be used to balance the tradeoff between exploration and exploitation and then tune the selection pressure. And further more,the orthogonal crossover without crossover rate is used instead of the binomial crossover,which can maintain the population diversity and accelerate the convergence rate. Experimental studies are carried out on a suite of 7 bound-constrained numerical benchmark functions. The results show that the proposed algorithm has better capability of maintaining the population diversity and faster convergence than the classical differential evolution and several classic differential evolution variants.
Bibliography:Ding Qingfeng , Qiu Xiang （School of Electrical and Automation Engineering, East China Jiaotong University, Nanchang 330013, P. R. China）
In order to reduce the pressure of parameter selection and avoid trapping into the local optimum,a novel differential evolution（ DE） algorithm without crossover rate is proposed. Through embedding cellular automata into the DE algorithm,those interactions among vectors are restricted within cellular structure of neighbors while the cell own evolution,which may be used to balance the tradeoff between exploration and exploitation and then tune the selection pressure. And further more,the orthogonal crossover without crossover rate is used instead of the binomial crossover,which can maintain the population diversity and accelerate the convergence rate. Experimental studies are carried out on a suite of 7 bound-constrained numerical benchmark functions. The results show that the proposed algorithm has better capability of maintaining the population diversity and faster convergence than the classical differential evolution and several classic differential evolution variants.
differential evolution（DE） cellular automata orthogonal crossover balancing tradeoff selective pressure
11-3683/N
ISSN:1006-6748
DOI:10.3772/j.issn.1006-6748.2017.04.012