Title: Improvement of Genetic Algorithm Performance for Identification of Cultivation Process Models
Keywords: Genetic Algorithm, Genetic Parameters, Identification, Cultivation Process
Abstract: In the genetic algorithms, there are many parameters and settings that can be implemented differently in various problems. There is no general theory about tuning the genetic algorithm parameters. In this paper some adjustments of genetic parameters, according to the identification problem of cultivation process models, are done to improve the genetic algorithm performance. Based on the proposed procedure for genetic parameters tuning and series of experiments a successful adjustment of the algorithm is done. The results show that the identification procedure leads to higher order of accuracy of the obtained parameter estimations and to 2.5 times decrease of the computation time. In order to validate the proposed genetic algorithm two experimental data sets of a recombinant E. coli BL21(DE3)pPhyt109 cultivation process are used. Based on the regarded genetic algorithm successful identification is done and fulfilled model verification shows the model adequateness.
Вид: публикация в национален форум