Article ID: | iaor20084669 |
Country: | Netherlands |
Volume: | 176 |
Issue: | 1 |
Start Page Number: | 60 |
End Page Number: | 76 |
Publication Date: | Jan 2007 |
Journal: | European Journal of Operational Research |
Authors: | Ali M.M., Kaelo P. |
Modifications in crossover rules and localization of searches are suggested to the real coded genetic algorithms for continuous global optimization. Central to our modifications is the integration of different crossover rules within the genetic algorithm. Numerical experiments using a set of 50 test problems indicate that the resulting algorithms are considerably better than the previous version considered and offer a reasonable alternative to many currently available global optimization algorithms, especially for problems requiring ‘direct search type’ methods.