Advertisement

Computer Science - Research and Development

, Volume 33, Issue 1–2, pp 233–234 | Cite as

Poster Abstract: Real-time load prediction with high velocity smart home data stream

  • Christoph Doblander
  • Martin Strohbach
  • Holger Ziekow
  • Hans-Arno Jacobsen
Special Issue Paper
  • 165 Downloads

Abstract

This poster addresses the use of smart home data to continuously predict the aggregated energy consumption of individual households. We introduce a device level energy consumption dataset recorded over 3 years wich includes high resolution energy measurements from electrical devices collected within a pilot program. Using data from that pilot, we analyze the performance of various machine learning mechanisms for continuous short-term load prediction.

Keywords

Electrical load prediction Machine learning 

Notes

Acknowledgements

This work has been supported in part by the Alexander von Humboldt foundation, the H2020 project HOBBIT under the Grant Agreement Number 688227 and by the German Federal Ministry of Economics and Technology, project PeerEnergyCloud which is part of the Trusted Cloud Program.

References

  1. 1.
    NY Times: solar experiment lets neighbors trade energy among themselves (2017). https://nyti.ms/2nhchnU
  2. 2.
    Siano P (2014) Demand response and smart gridsa survey. Renew Sustain Energy Rev 30:461–478Google Scholar
  3. 3.
    Doblander et al. Real-time load prediction with high velocity smart home data stream. arXiv:1708.04613
  4. 4.
    Veit A et al (2014) Household electricity demand forecasting: benchmarking state-of-the-art methods. In: Proceedings of the 5th international conference on future energy systems, e-Energy ’14. ACM, New YorkGoogle Scholar
  5. 5.
    Pedregosa F et al (2011) Scikit-learn: machine learning in Python. J Mach Learn Res 12:2825–2830MathSciNetzbMATHGoogle Scholar
  6. 6.
    Hyndman RJ, Koehler AB (2006) Another look at measures of forecast accuracy. Int J Forecast 22(4):679–688CrossRefGoogle Scholar
  7. 7.
    Edwards RE, New J, Parker LE (2012) Predicting future hourly residential electrical consumption: a machine learning case study. Energy Build 49:591–603CrossRefGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany 2017

Authors and Affiliations

  • Christoph Doblander
    • 1
  • Martin Strohbach
    • 2
  • Holger Ziekow
    • 3
  • Hans-Arno Jacobsen
    • 1
  1. 1.TU-München - InformatikMünchenGermany
  2. 2.AGT Group (R&D) GmbHDarmstadtGermany
  3. 3.HS-Furtwangen - Fakultät WirtschaftsinformatikFurtwangenGermany

Personalised recommendations