Loading…
An introduction of dominant genes in genetic algorithm for FMS
This paper proposes a new idea, namely genetic algorithms with dominant genes (GADG) in order to deal with FMS scheduling problems with alternative production routing. In the traditional genetic algorithm (GA) approach, crossover and mutation rates should be pre-defined. However, different rates app...
Saved in:
Published in: | International journal of production research 2008-08, Vol.46 (16), p.4369-4389 |
---|---|
Main Authors: | , , |
Format: | Article |
Language: | English |
Subjects: | |
Citations: | Items that this one cites Items that cite this one |
Online Access: | Get full text |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | This paper proposes a new idea, namely genetic algorithms with dominant genes (GADG) in order to deal with FMS scheduling problems with alternative production routing. In the traditional genetic algorithm (GA) approach, crossover and mutation rates should be pre-defined. However, different rates applied in different problems will directly influence the performance of genetic search. Determination of optimal rates in every run is time-consuming and not practical in reality due to the infinite number of possible combinations. In addition, this crossover rate governs the number of genes to be selected to undergo crossover, and this selection process is totally arbitrary. The selected genes may not represent the potential critical structure of the chromosome. To tackle this problem, GADG is proposed. This approach does not require a defined crossover rate, and the proposed similarity operator eliminates the determination of the mutation rate. This idea helps reduce the computational time remarkably and improve the performance of genetic search. The proposed GADG will identify and record the best genes and structure of each chromosome. A new crossover mechanism is designed to ensure the best genes and structures to undergo crossover. The performance of the proposed GADG is testified by comparing it with other existing methodologies, and the results show that it outperforms other approaches. |
---|---|
ISSN: | 0020-7543 1366-588X |
DOI: | 10.1080/00207540600632190 |