Automatic Configuration of Hybrid Mathematical Models

Schumann, René (University of Applied Sciences and Arts Western Switzerland (HES-SO Valais-Wallis)) ; Barry, Michael (University of Applied Sciences and Arts Western Switzerland (HES-SO Valais-Wallis))

In this paper we address the problems faced for automatic configuration of flexible hybrid models. We approach this problem with concepts from the computer science field including object orientated design and dependency injection to create a new Hybrid Model that combines traditional modelling methods with a flexible design. This type of model is able to drastically change its functional behaviour, allowing it to simulate a larger variety of scenarios of varying complexity. We then use AI methods to automatically configure the model to reduce its complexity to the minimum while having minimal impact on the models accuracy. A small example is demonstrated where this method is used to configure the market environment for a hydro-power plant model, allowing us to determine which set of markets are most profitable for any given plant configuration . Furthermore, the use of flexible hybrid models opens up the possibility for further AI methods to be used in conjunction with mathematical models.


Conference Type:
full paper
Faculty:
Economie et Services
School:
HEG-VS
Institute:
Institut Informatique de gestion
Subject(s):
Informatique
Publisher:
Toulouse, France, 5 September 2016
Date:
Toulouse, France
5 September 2016
2016
Pagination:
6 p.
Published in:
Proceedings of the 18th International Configuration Workshop within CP 2016 Conference
Appears in Collection:

Note: The status of this file is: restricted


 Record created 2016-11-29, last modified 2019-06-11

Fulltext:
Download fulltext
PDF

Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)