Optimum tuning parameters for Encapsulated Evolution Strategies : Results for a nonlinear regression problem
Loading...
Date
1998-11-08
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Universität Dortmund
Abstract
The prediction of certain thermodynamic properties of pure substances and mixtures with calculation methods is a frequent task during the process design in chemical engineering. Group contribution models divide the molecules into functional groups and if the model parameters for theses groups are known, predictions of compounds that comprise these groups are possible. The model parameters have to be fitted to experimental data, which leads to a multi-parameter multimodal optimization problem. In this paper the optimization of the tuning parameters of Evolution Strategies and different methods of parameter fitting regarding the number of parameters are presented.