Adaptively exploring the feature space of flowsheets
- Johannes Höller,
- Martin Bubel,
- Raoul Heese,
- Patrick Ludl,
- Patrick Schwartz,
- Jan Schwientek,
- Norbert Asprion,
- Martin Wlotzka,
- Michael Bortz
Johannes Höller
Fraunhofer Institute for Industrial Mathematics ITWM
Author ProfileMartin Bubel
Fraunhofer Institute for Industrial Mathematics ITWM
Author ProfileRaoul Heese
Fraunhofer Institute for Industrial Mathematics ITWM
Author ProfilePatrick Ludl
Fraunhofer Institute for Industrial Mathematics ITWM
Author ProfilePatrick Schwartz
Fraunhofer Institute for Industrial Mathematics ITWM
Author ProfileJan Schwientek
Fraunhofer Institute for Industrial Mathematics ITWM
Author ProfileMichael Bortz
Fraunhofer Institute for Industrial Mathematics ITWM
Corresponding Author:michael.bortz@itwm.fraunhofer.de
Author ProfileAbstract
Simulation and optimization of chemical flowsheets rely on the solution
of a large number of non-linear equations. Finding such solutions can be
supported by constructing machine-learning based surrogates, relating
features and outputs by simple, explicit functions. In order to generate
training data for those surrogates computationally efficiently, schemes
to adaptively sample the feature space are mandatory. In this article,
we present a novel family of utility functions to favor an adaptive,
Bayesian exploration of the feature space in order to identify regions
that are convergent, fulfill customized inequality constraints and are
Pareto-optimal with respect to conflicting objectives. The benefit is
illustrated by small toy-examples as well as by industrially relevant
chemical flowsheets.