Advertisement

Power Prediction in Smart Grids with Evolutionary Local Kernel Regression

  • Oliver Kramer
  • Benjamin Satzger
  • Jörg Lässig
Part of the Lecture Notes in Computer Science book series (LNCS, volume 6076)

Abstract

Electric grids are moving from a centralized single supply chain towards a decentralized bidirectional grid of suppliers and consumers in an uncertain and dynamic scenario. Soon, the growing smart meter infrastructure will allow the collection of terabytes of detailed data about the grid condition, e.g., the state of renewable electric energy producers or the power consumption of millions of private customers, in very short time steps. For reliable prediction strong and fast regression methods are necessary that are able to cope with these challenges. In this paper we introduce a novel regression technique, i.e., evolutionary local kernel regression, a kernel regression variant based on local Nadaraya-Watson estimators with independent bandwidths distributed in data space. The model is regularized with the CMA-ES, a stochastic non-convex optimization method. We experimentally analyze the load forecast behavior on real power consumption data. The proposed method is easily parallelizable, and therefore well appropriate for large-scale scenarios in smart grids.

Keywords

Local Model Smart Grid Power Demand Power Prediction Kernel Regression 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Download centre Eirgrid, http://www.eirgrid.com/
  2. 2.
    Agarwal, V., Bougaev, A., Tsoukalas, L.H.: Kernel regression based short-term load forecasting. In: ICANN, vol. (2), pp. 701–708 (2006)Google Scholar
  3. 3.
    Alfares, H., Nazeeruddin, M.: Electric load forecasting: literature survey and classifcation of methods. International Journal of Systems Science 33(1), 23–34 (2002)CrossRefzbMATHGoogle Scholar
  4. 4.
    Beyer, H.-G., Schwefel, H.-P.: Evolution strategies - A comprehensive introduction. Natural Computing 1, 3–52 (2002)MathSciNetCrossRefzbMATHGoogle Scholar
  5. 5.
    Clark, R.: A calibration curve for radiocarbon dates. Antiquity 46(196), 251–266 (1975)CrossRefGoogle Scholar
  6. 6.
    Hansen, N.: The CMA evolution strategy: A tutorial. Technical report, TU Berlin, ETH Zürich (2005)Google Scholar
  7. 7.
    Lora, A.T., Santos, J.M.R., Expósito, A.G., Ramos, J.L.M., Santos, J.C.R.: Electricity market price forecasting based on weighted nearest neighbors techniques. IEEE Transactions on Power Systems 22(3), 1294–1301 (2007)CrossRefGoogle Scholar
  8. 8.
    Nadaraya, E.: On estimating regression. Theory of Probability and Its Application 10, 186–190 (1964)CrossRefzbMATHGoogle Scholar
  9. 9.
    Nogales, F.J., Contreras, J., Conejo, A.J., Espinola, R.: Forecasting next-day electricity prices by time series models. IEEE Transactions on Power Systems 17, 342–348 (2002)CrossRefGoogle Scholar
  10. 10.
    Ostermeier, A., Gawelczyk, A., Hansen, N.: A derandomized approach to self-adaptation of evolution strategies. Evol. Comput. 2(4), 369–380 (1994)CrossRefGoogle Scholar
  11. 11.
    Rumelhart, D., Hintont, G., Williams, R.: Learning representations by back-propagating errors. Nature 323(6088), 533–536 (1986)CrossRefGoogle Scholar
  12. 12.
    Watson, G.: Smooth regression analysis. Sankhya Series A 26, 359–372 (1964)MathSciNetzbMATHGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2010

Authors and Affiliations

  • Oliver Kramer
    • 1
  • Benjamin Satzger
    • 1
  • Jörg Lässig
    • 1
  1. 1.International Computer Science InstituteBerkeleyUSA

Personalised recommendations