Advertisement

Analysis and Classification of Energy Requirement Situations Using Kohonen Feature Maps within a Forecasting System

  • Steffen Heine
  • Ingo Neumann
Part of the Advances in Industrial Control book series (AIC)

Abstract

Improving the accuracy of the electrical energy demand forecasting in the short term range up to seven days can decrease the operation costs of the energy system significantly by the following optimising of the energy management. Because of the different objectives of energy management in different energy systems a considerable amount of engineering power is necessary to build a forecast model for each application case. The efficiency of using the Kohonen Feature Map for a stepwise automatisation of this process is discussed by means of two application cases. Finally, the embedding of the new analysis capability in a modular-constructed forecasting system is presented.

Keywords

Artificial Neural Network Input Space Forecast System Unit Commitment Load Demand 
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.
    Bezdek, J. C.: Pattern Recognition with Fuzzy Objective Function Algorithms. Plenum, New York, 1981zbMATHGoogle Scholar
  2. 2.
    Gottschalk, H., Heine, S., Fox, B., Neumann, I.: Economic Operation of a Power System with a Significant Amount of Controllable Load. Proceedings of 29th Universities Power Engineering Conference UPEC, University of Galway, 673–675, September 1994Google Scholar
  3. 3.
    Heine, S., Neumann, I.: Information Systems for Load-Data Analysis and Load Forecast by Means of Specialised Neural Nets. Proceedings of 28th Universities Power Engineering Conference UPEC, Staffordshire University, 279–282, Sept. 1993Google Scholar
  4. 4.
    Heine, S., Wilde, S.: Online-Prognose des Elektroenergieverbrauches eines sächsischen Stadtwerkes. Report, TH Leipzig, August 1994Google Scholar
  5. 5.
    Heine, S., Neumann, I.: Optimizing Load Forecast Models Using an Evolutionary Algorithm. Second European Congress on Intelligent Techniques and Soft Computing (EUFIT), 1690–1694, Aachen, September 1994Google Scholar
  6. 6.
    Kohonen, T.: Learning Vector Quantisation and the Self Organising Map. Theory and Applications of Neural Networks, 235–242, Springer Verlag, London, Berlin, Heidelberg, 1989Google Scholar
  7. 7.
    Macabrey, N., Baumann, T., Germond, A. J.: Prévision de charge dans un réseau électrique à laide du réseau de neurones de Kohonen. Bulletin SEV/VSE 83 (1992) 5, 13–19Google Scholar
  8. 8.
    Schreiber, H., Heine, S.: Einsatz von Neuro-Fuzzy-Technologien für die Prognose des Elektroenergieverbrauches an “besonderen” Tagen. Proceedings of the 4th Fuzzy Days, Universität Dortmund, 274–281, Juni 1994Google Scholar
  9. 9.
    Ultsch, A.: Konnektionistische Modelle und ihre Integration mit wissensbasierten Systemen. Research Report Nr. 396, Universität Dortmund, January 1991Google Scholar

Copyright information

© Springer-Verlag London Limited 1995

Authors and Affiliations

  • Steffen Heine
    • 1
  • Ingo Neumann
    • 1
  1. 1.Best Data Engineering GmbH BerlinBerlinGermany

Personalised recommendations