Master Thesis Meta-learning for Fast Identification of Models of Electric Machines

The identification of accurate simulation models of electric machines is a crucial step for the design of high-performing controllers, fault diagnosis, and many other tasks. Often, the time available for performing measurements on a physical device is limited. One general approach to still obtain an accurate model under such constraints is transfer- and meta-learning. We further aim to combine this with knowledge on the physics of the electric machine by forming a grey-box model, which promises to further reduce the required measurement time. The goal of this thesis is to investigate grey-box meta-learning approaches for accurate identification of models of electric machines, given tight measurement time constraints.

Responsibilities include:

• During your thesis you will conduct comprehensive literature research on meta-learning and transfer-learning approaches.
• You will familiarize yourself with physical models of electric machines.
• Furthermore, you will design and implement meta-learning models, as well as train algorithms for the identification of electric machines.
• Finally, you will evaluate and benchmark the developed approaches through simulation.

Your profile

• Education: studies in the field of Mathematics, Physics, Electrical Engineering, Cybernetics, Computer Science or comparable
• Experience and Knowledge: in modelling of dynamical systems, machine learning and Python
• Personality and Working Practice: you excel at working independently and systematically organizing your tasks
• Languages: very good in English

Contact & additional information

Start: according to prior agreement

Duration: 6 months

Requirement for this thesis is the enrollment at university. Please attach your CV, transcript of records, examination regulations and if indicated a valid work and residence permit.