InterCriteria Analysis Approach for Comparison of Simple and Multi-population Genetic Algorithms Performance

Abstract

Intercriteria analysis approach, based on the apparatuses of index matrices and intuitionistic fuzzy sets, has been here applied for comparison of simple and multi-population genetic algorithms performance. Six kinds of simple genetic algorithms and six kinds of multi-population genetic algorithms, differing in the execution order of the main genetic operators selection, crossover and mutation are in the focus of current investigation. Intercriteria analysis approach is implemented to assess the performance of mentioned above genetic algorithms for the purposes of parameter identification of Saccharomyces cerevisiae fed-batch fermentation process. For the completeness, the performance of altogether twelve algorithms have been also assessed. Degrees of agreement and disagreement between the algorithms outcomes, namely convergence time and model accuracy, from one hand, and model parameters estimations, from the other hand, have been established. The obtained results after the application of intercriteria analysis have been compared and outlined relations have been thoroughly discussed.