Graduate School

Project A15: Online State Tracking with Data-Driven Process Models

Person in Charge: Dipl.-Inform. Susanne Fischer
Motivation and Objectives:
The goal of this project is the optimization of the process and the process chain, respectively. For this reason, process information and the process dynamic will be modeled in real-time and automated by using simulation data and experiments.
A knowledge-based modeling of the dynamic process is intended to create an interpretable result. This will be used for observation and prediction of process features depending on the input parameters.
Research Results
  • Luenberger observer
  • simultaneous symbolic regression and parameter estimation in order to determine the model structure and parameter values
  • methods to derive symbolic modeling elements for the relation between
    • measured variables and state variables, as well as
    • change of state and control parameters
  • Description of the dynamic of the process
  • evaluation by reference to the observed process chain
  • Reduction of high-dimensional data to a few features to describe the change of the state during a deep drawing process