Abstract : The general flowshop scheduling problem is approximately solved using a genetic algorithm. Chromosomes are classically encoded using permutation strings for each machine. The data-oriented operators use both the permutation strings and problem instances. The beginning of one parent is copied inside the first offspring which is completed with a priority rule generator using data of the second parents. The second offspring is obtained in a similar way. Several cross-over operators are designed. Some of them work only on the permutation encoding, other ones are data oriented.
https://hal.inria.fr/inria-00098866
Contributeur : Publications Loria
<>
Soumis le : mardi 26 septembre 2006 - 08:39:37
Dernière modification le : jeudi 11 janvier 2018 - 06:19:48