 |
The research project
ACON is funded by
the german BMBF
»Bundesministerium für Bildung, Wissenschaft, Forschung und
Technologie« under grant 01IN510C8. |
BMBF Joint project ACON
Adaptive Modelling for Optimization, Planing and Control
The common goal of the joint project was the development and
application of methods and techniques from the fields of Computational
Intelligence and Distributed Artificial Intelligence to problems that
arise in the area of adaptive control of processes. The focus of the
research group at Institut für Informatik lay in the modeling and
identification of nonlinear processes using neural networks. The
achieved results can be categorized into four topic groups. Two of
them, namely the identification of time delays and hysteresis, deal
with special kinds of isolated, nonlinear dynamical systems, which
often appear in technical systems. The other two deal with the
integration of component models to entire system models: In one case
based on the use of local models for different process state space
areas, in the other one by integration of models corresponding to the
actual physical structure of the system.
The ACON-Project ended 31.12.1999.
The final report of this project was published as
Technical Report FKI-239-00 (only in german).
Publications:
- Ungerer, C. (2000): Identifying time delays using neural networks.
In Dimitris Tsaptsinos, editor, Engineering Problems - Neural Network Solutions,
Proceedings of the International Conference on Engineering Applications of
Neural Networks (EANN2000).
- Ungerer, C., Stübener, D., Kirchmair, and Sturm, M. (1999):
Supporting Traditional Controllers of Combustion Engines by
means of Neural Networks.
In Bernd Reusch, editor, Computational Intelligence -
Theory and Applications, International conference
6th Fuzzy Days 1999. Springer.
- Ungerer, C. (1999):
Identifying Time-Delays in
Nonlinear Control Systems: An Application of the Adaptive Time-Delay Neural Network.
In Masoud Mohammadian, editor, Computational Intelligence for Modelling, Control, and Automation
(CIMCA '99).
IOS Press.
- Kirchmair, C. (1999): Ein Gradientenabstiegsverfahren zum Schätzen der
Parameter des Preisach Modells für Hysterese.
Forschungsberichte Künstliche Intelligenz FKI-231-99, Institut für Informatik,
Technische Universität München.
- Eder et al (1998): Adaptive Modellbildung für Anwendungen in der
Fahrzeugentwicklung.
In Statustagung des BMBF: Intelligente Systeme.
DLR Informationsverarbeitung.
- Brychcy, T. (1998): Prestructured recurrent neural networks.
In Brauer, W.(Hrsg.), Fuzzy-Neuro Systems `98 – Computational Intelligence.
Proceedings in Artificial Intelligence, S. 210-217. Infix Verlag, St. Augustin.
- Ungerer, C. (1998):Neuronale Modellierung von Totzeiten in nichtlinearen dynamischen Systemen.
Diplomarbeit, Institut für Informatik,
Technische Universität München.
- Sturm, M., Brychcy, T., and Kirchmair, C. (1997):
AMoC - The ACON Model
Classes.
Forschungsberichte Künstliche Intelligenz
FKI-224-97. Institut für Informatik, Technische
Universität München
- Sturm, M., Eder, K., Brauer, W. and Gonzáles, J. C. (1997):
Hybridization of Neural
and Fuzzy Systems by a Multi Agent Architecture for
Motor Gearbox Control.
Fuzzy Sets and Systems
89(3), pp. 309-319
- Sturm, M., and Brychcy, T. (1997):
On-Line
Prozeßraumkartierung mit ellipsoider Vektorquantisierung
zur lokalen Modellbildung.
In Fuzzy-Neuro-Systeme ´97 -
Computational Intelligence, Proceedings in Artificial
Intelligence,
pp. 463-470. A. Grauel, W.Becker and F.Belli
(Edt.). Infix Verlag, St. Augustin.
- Sturm, M., and Eder, K. (1996):
Self-organizing Process
State Detection for On-line Adaptation Tasks.
In
Solving Engineering Problems with Neural Networks,
Proceedings of the International Conference
EANN´96,
pp. 33-36. A. B. Bulsari, S. Kallio
and D. Tsaptsinos (Edt.). Systeemitekniikan seura ry,
London.
Last modified: Thu Jan 17 12:59:52 2002
