english  | Home | Impressum | KIT

Machine learning applied to forecasting of the maximum current capacity of electrical networks

Machine learning applied to forecasting of the maximum current capacity of electrical networks
Typ:Bachelor-/ Masterarbeit
Datum:offen (zu vergeben)
Betreuer:

M. Eng. Gabriela Molinar

Machine learning applied to forecasting of the maximum current capacity of electrical networks

.
.
.
.

Background

The constant increase of wind parks in north Germany and the fact that the greatest consumption of energy is in the south of the country mean the need to carry more energy over hundreds of kilometers. This leads to either the construction of more transmission lines or a better use of the existent electrical network.
With overhead line monitoring systems, the already installed transmission lines can be used more efficiently and even a prediction of the maximum current capacity of the network can be done. Knowing this, the Transmission System Operators (TSOs) can estimate better the use of the network depending on the variable generation of renewable energy sources.
Machine learning is a promising tool to do forecasting. Your work would be to play with it to see what kind of algorithm fits better our goals. 

 

Tasks

  • Selection of a machine learning algorithm of your preference.
  • Implementation of the algorithm.
  • Tests with real data.
  • Analysis of the results: A good or bad fit of the forecasted data is always useful to understand better our system. 

 

Required skills

  • Motivation and creativity!
  • Student of electrical engineering.