Energy peak load prediction

  • Sponsored: BMW AG
  • Scientific Lead: Dominik Schniertshauer, Benny Kneissl, Vadim Gorski (co-mentor)
  • Project Lead: Dr. Ricardo Acevedo Cabra
  • Semester: Winter 2017/2018

Every year BMW Group is spending millions for energy used in BMW Group production plants. Since 2006 the energy consumed could be reduced significantly while the production numbers increased. The energy cost is typically divided in two parts: first, the whole sum of the energy consumption in a year and, second, the maximum energy peak load, i.e. the maximum energy usage within 15 minutes in a year. The goal of this project is to predict these energy peak loads based on the planned production program with a high accuracy. This would help to switch for these time spans to internal energy resources and avoid the peak loads regarding foreign feed-ins. Hence, the application of an accurate prediction model in real-time will have a direct influence on the energy costs.

The results of this project were summarized in the slides of the final presentation and fully explained in the final documentation.