Below, you find a list with topics of open or running theses and student projects.
Click on the boxes to obtain more information about the various topics. In case you are interested in one of the proposed projects, please contact the listed supervisor via email and attach a recent transcript of records. Note that the documentation of each project can be done in German or English. Further note that we are always open to own suggestions of control-related theses topics. This also includes projects with industrial partners. However, external theses require clear problem statements and profound methodology.
Thesis or Project: Neural network-based predictive control
Neural networks (NN) are very popular in many fields due to their ability to approximate a large class of functions with arbitrary accuracy. For piecewise affine predictive control schemes, even exact fits are possible as studied in this project.
Thesis or Project: Privacy-preserving distributed estimation
Clock synchronization in networks or position estimation in robot swarms can be realized with affine averaging. However, this distributed estimation algorithm requires the exchange of sensitive data. The project addresses a secure implementation.
Thesis or Project: Encrypted control for secure automation
Modern control systems tend to be networked and distributed. While these systems offer exciting new possibilities, they also involve privacy and security concerns. Encrypted control tackles this issue.
Project: Illustrating MIMO control with the Quanser AERO
With its individually controllable propellers, the Quanser AERO provides the perfect platform for the illustration of advanced control concepts for multi-input multi-output (MIMO) systems.
Thesis or Project: Control concepts for automated trading
Optimal decision-making under stochastic influences is a control problem at its heart and highly relevant in many disciplines such as trading, production, transportation and logistics.
Thesis: Development of a control approach for multi-stage press hardening
Until now, multi-stage press hardening has been limited to special applications due to the difficult-to-predict thermomechanical interactions, which will be overcome by implementing a closed-loop control.