A New Differential Evolution Algorithm with Random Parameters

Differential evolution algorithm’s performance often depends heavily on the parameter settings. Based on analyzing the influence of the parameters setting in the experiment, the effects and the optimal selection of those major parameters on DE are analyzed, and some conclusions are derived. A new di...

Full description

Saved in:
Bibliographic Details
Published inApplied Mechanics and Materials Vol. 556-562; pp. 3614 - 3621
Main Author Huang, Shao Rong
Format Journal Article
LanguageEnglish
Published Zurich Trans Tech Publications Ltd 01.05.2014
Subjects
Algorithm Improvement
Parameter Analysis
Numerical Optimization
Differential Evolution (DE) Algorithm
Online AccessGet full text

Cover

More Information
Summary:Differential evolution algorithm’s performance often depends heavily on the parameter settings. Based on analyzing the influence of the parameters setting in the experiment, the effects and the optimal selection of those major parameters on DE are analyzed, and some conclusions are derived. A new differential evolution algorithm which the scale constant (F) and crossover constant (CR) are generated as random numbers within a certain range in each iteration process is proposed. The experimental results shows that the new algorithm is simple, easy to realize and can get higher precision and better stability.
Bibliography:Selected, peer reviewed papers from the 2014 International Conference on Mechatronics Engineering and Computing Technology (ICMECT 2014), April 9-10, 2014, Shanghai, China
ISBN:3038351156
9783038351153
ISSN:1660-9336
1662-7482
1662-7482
DOI:10.4028/www.scientific.net/AMM.556-562.3614