AUTOtemp

Autonomes Temperierungsverfahren für den Metall-Druckguss

The setup of temperature control systems in metal die casting is time-consuming and error-prone, since it is currently based on intuition, 'trial and error' and experience. Due to increasing complexity and higher quality requirements, errors are hardly avoidable and the setup process can often take several days. We therefore plan to develop a measurement and control system for temperature control which is supported by several novel machine learning approaches. While conventional machine learning methods usually require far too long training times and are hardly operable for a person not familiar with the subject, several novel approaches considered in the project enable a high prediction quality even with small training groups and is highly automated. Thereby, the system becomes manageable even for machine operators without a background in statistics.

 

Project duration: 01.02.2019 - 31.01.2021
Project lead: Prof. Dr. rer. nat. Johannes Gottschling
Project staff: Saad Alvi, M.Sc.
Christian Noss, M.Sc.
Consortium: thermobiehl Apparatebau GmbH
Universität Duisburg-Essen

 

logo_BMWilogo_ZIM ​​

University of Duisburg-Essen Institute for Technologies of Metals

Address
Institut für Technologien der Metalle
Friedrich-Ebert-Str. 12
47119 Duisburg / Laar

Postal address
Universität Duisburg-Essen
Campus Duisburg
Institut für Technologien der Metalle
Forsthausweg 2
47057 Duisburg