Memetic Algorithm with Population Management (MA|PM) for Multi-objective Network Design.

No Thumbnail Available

Date

2014

Journal Title

Journal ISSN

Volume Title

Publisher

The Computer Chapter of the Institute of Electrical & Electronics Engineers (IEEE), Nigeria.

Abstract

A Memetic Algorithm with Population Management (MA|PM) is employed to solve Multi-objective Network Design Problem. The algorithm was tested with three randomly generated networks of varying sizes. Results obtained were compared with the results obtained when MA was used. Overall, it was observed that MA|PM outperformed MA in terms of efficiency (computation time) for the three test problems. In addition, the results’ quality of MA|PM is superior to that of MA for 10-node network problem while it is inferior to that of MA for 36-node network problem. The results’ qualities of MA|PM and MA rank the same for 21-node network problem. The implication of these results is that MA|PM is always more efficient that MA regardless of problem size. However, the impact of population management on the effectiveness of MA is inversely proportional to the size of the problem.

Description

Keywords

Memetic Algorithm,, Population Management, Efficiency

Citation

Collections