A Hybrid Differential Evolution Algorithm for Binary CSPs

Abstract:

Article Preview

A novel hybrid elements exchange/electromagnetism meta-heuristic differential evolution algorithm, named EEMDE, is proposed in this paper, avoiding the premature convergence of original DE algorithm. A metric to measure the Simplification of force exerted on a point is defined as the mutation rate F in the EEMDE, which is used to get an adaptive adjustment of F. EEMDE may produce slight disturbance on the original vector for enhancing the exploring capacity and avoid the DE to the "uphill" in the wrong direction forward. Experiments demonstrate that the convergence of EEMDE is faster than DE and simulations based on some CSPs express the effectiveness, efficiency and robustness of it.

Info:

Periodical:

Advanced Materials Research (Volumes 108-111)

Edited by:

Yanwen Wu

Pages:

328-334

DOI:

10.4028/www.scientific.net/AMR.108-111.328

Citation:

H. J. Fu "A Hybrid Differential Evolution Algorithm for Binary CSPs", Advanced Materials Research, Vols. 108-111, pp. 328-334, 2010

Online since:

May 2010

Authors:

Export:

Price:

$35.00

In order to see related information, you need to Login.

In order to see related information, you need to Login.