FH Bielefeld
University of
Applied Sciences

Overview of topics for theses

Our competencies

  • Decentralised energy systems
  • Energy storage devices and fuel cells
  • Electrical grids
  • Electrical engineering

Open positions and topics for theses and project papers

AGNES offers students the opportunity to work on student research, final theses and project work in their subject areas. In this, we strive to meet the students’ interests and possibilities. The topics can be adapted according to the extent of a project work, a bachelor thesis or a master thesis. Current vacancies for theses and project works are listed below.

Subject area “Predictive Systems”

This area covers the use of artificial intelligence in the electrical grid, in particular the prediction of volatile renewable energies such as PV plants or consumers. There are various topics that can be worked on in a project or thesis. Please find some examples of current project works and theses below:

  • Project work: Creation of a mathematical PV model for converting the solar radiation into the power generated by the PV plant and validation of the model with real measurement data
  • Thesis or project work: Self-learning artificial neural networks to minimise the predictive error in forecasting the PV performance of a plant when continuously learning current measurements
  • Thesis or project work: Forecasting individual household loads

If you are interested, please contact Katrin Schulte (katrin.schulte@fh-bielefeld.de)

Subject area “Control System of Low-Voltage Grids”

The electrical grid is a complex and dynamic system. New high-power consumers, such as electric vehicle charging stations, heat pumps or decentralized renewable energy systems, can in future endanger grid stability. These systems are mainly installed on low-voltage level. Historically, the low-voltage level is not monitored metrologically. In order to improve grid stability, the new consumers and renewable energies described above must be synchronized and controlled in an informative manner and the low-voltage level must become more observable. For this purpose, artificial neural networks, agent-based control algorithms and reinforcement learning are increasingly being used in this field. Matlab/Simulink and Python are mainly used for this. Basic knowledge of these is desirable, but not necessary. The following topics are suitable for project works and theses:

  • Modelling and simulation of detailed generation and consumption systems in the electrical low-voltage grid
  • Real-time simulation of electrical grids on a hardware grid simulator (Opal-RT)
  • Coupling of hardware models of electrical systems with models of electrical grids in software Hardware in the loop (HIL) and power hardware in the loop (PHIL)
  • Classical status analysis of electrical grids adapted to distribution grids or status analysis by means of artificial neural networks
  • Definition of reinforcement learning algorithms (Q-learning/deep Q-learning) for autonomous control systems of low-voltage grids

If you are interested, please contact Michael Kelker (michael.kelker@fh-bielefeld.de)