Predictive Control of Waste Incinerator

Project Members:

  • apl. Prof. Dr. Matthias Reuter (Leader)
  • Dipl.-Inf. Sven Birkenfeld (Project Staff)

Partner:

  • Dr.-Ing. Stefan Vodegel, CUTEC-Institut GmbH, Clausthal, Germany

Funding: AIF (Arbeitsgemeinschaft industrieller Forschungsvereinigungen "Otto von Guericke" e.V.), 174.000 € (of 360.000 €)

Duration: 01/2006 - 12/2008

Project Description: With help of neural nets, a waste incinerator is controlled in a predictive way. As the implicit model of the waste incinerator we use a closed SOM, whose activities are decoded by a downstream feed-forward net. In consequence, the predictive values of the control can be calculated. This is a new approach in the area modelling and simulation; in particular, it is the first industrial application of the theory of Computing with Activities.

Contact E-Mail: matthias.reuter@tu-clausthal.de