Control Engineering
In the field of control engineering, the first focus is on the development of data-driven methods. Process model-based analysis and synthesis methods are only used in selected cases in industrial practice because of the high effort involved in modeling. The Department of Measurement and Control Engineering researches and develops methods to ideally be able to carry out the process of non-linear system identification fully automatically using computational intelligence (CI) methods as well as conventional statistical and numerical methods ("push-button model"). On the one hand, tasks for which prior knowledge is usually required (such as structural decisions) should be completed automatically. On the other hand, the efficiency and effectiveness of the methods developed should be improved. Area-wise affine and Takagi-Sugeno models are of particular interest. The models are developed with regard to their intended use, with particular focus on non-linear control design, error diagnosis, real-time simulation and prognosis in the specialist area.
As a further focus, the area of complex systems is being developed in the sense of large coupled systems. Conventional methods of control engineering and system identification are impractical for such systems. On the one hand, the department of measurement and control technology researches and develops approximate/qualitative CI methods in order to be able to dynamically describe, analyze and influence such systems with little modeling effort. Other work addresses methods for efficiently and effectively examining large amounts of metric data for analysis and modeling.
The application focus of the work is in the areas of mechatronics, automotive and production systems.
Completed Projects
- EEpBeton
- Digital Twin of Injection Molding (DIM)
- Automatische Kalibrierung komplexer Sensorsysteme am Beispiel mikromagnetischer Materialcharakterisierungsverfahren
- Fault detection and isolation in process plants using Probabilistic Graphical Models
- Prognose des Randschichtzustandes für die robuste Regelung eines Drehprozesses unter Einsatz von in-process Messtechnik und datengetriebener Softsensorik
- Digitalization in Material Technology
- Experiment Design for the Identification of locally affine multi models
- Methods for automated model structure selection for the identification of dynamic Takagi-Sugeno fuzzy models
- Methods for automated model structure selection for the identification of dynamic Takagi-Sugeno fuzzy models
- On identification of mechatronic actuators with friction in motor vehicles
- Methods for task allocation and cooperation in mobile multi-robot systems
- On identification of mechatronic actuators with friction in motor vehicles
- Classification-based Online Fault Recognition
- Analysis of large and coupled systems with methodes of complex networks
- Fuzzy modeling of nonlinear uncertain dynamic systems
- Methods for dynamic modeling of nonlinear mechatronic systems
- Early detection and decision support for critical situations in the production environment