ETI 03-20 Bachelor- or Master Thesis : Development and application of a machine learning based methodology for the prediction of photovoltaic power
Karlsruhe Institute of Technology (KIT) – The Research University in the Helmholtz Association creates and imparts knowledge for the society and the environment. It is our goal to make significant contributions to mastering the global challenges of mankind in the fields of energy, mobility, and information. For this, about 9300 employees of KIT cooperate in a broad range of disciplines in research, academic education, and innovation.
Institute of Electrical Engineering (ETI)
Global warming and the exhaustion of fossil fuels are increasing the significance of energy supply alternatives. As a renewable energy source, photovoltaics offers an attractive solution for generating clean energy. However, during the day, the power generated by photovoltaic systems fluctuates due to the position of the sun and changing weather conditions. The partial unpredictability of the available power is a challenge for energy network operators, which is being addressed by optimizing prediction methods, among other things. In recent years, the trend towards data analysis and machine learning has increased in many application areas, including photovoltaic power prediction. The photovoltaic field of KIT Campus North provides an attractive data basis for the development of a power prediction.
Within the framework of a thesis, a methodology for the prediction of photovoltaic power is to be developed and tested on the basis of the current state of research. As a basis, already existing prediction models at the institute as well as an extensive database of historical data will be used.
The scope of the task will be worked out individually, the rough steps are as follows. First of all, the state of research is to be determined by means of a literature search and on this basis a suitable methodology is to be selected. Subsequently, the most important parameters are to be identified and their selection optimised if necessary. With the elaborated basis a corresponding photovoltaic power prediction shall be developed and trained. By means of common evaluation methods the results shall be validated and statements about the methodology shall be made.
as soon as possible
You have programming skills (ideally in Matlab, R and Python), experience in data analysis and machine learning, reliability, an independent way of working as well as a fast comprehension and very good German or English skills.
Field of study: electrical engineering, mechanical engineering, mathematics, process engineering
limited to 3-6 months.
Contact person in line-management
Fr. Anna Starosta, email@example.com
Tel.: 0721 / 608 28380
If qualified, severely disabled persons will be preferred.
Please apply online using the button below for this vacancy number ETI 03-20.
Personnel Support is provided by
Telefon: +49 721 608-25184,
Hermann-von-Helmholtz-Platz 1, 76344 Eggenstein-Leopoldshafen, Germany