Worst-Case Modelling for Management Decisions under Incomplete Information, with Application to Electricity Spot Markets

  • Mercedes Esteban-Bravo
  • Berc Rustem
Conference paper

DOI: 10.1007/3-540-36626-1_2

Volume 9 of the book series Advances in Computational Management Science (AICM)
Cite this paper as:
Esteban-Bravo M., Rustem B. (2007) Worst-Case Modelling for Management Decisions under Incomplete Information, with Application to Electricity Spot Markets. In: Kontoghiorghes E.J., Gatu C. (eds) Optimisation, Econometric and Financial Analysis. Advances in Computational Management Science, vol 9. Springer, Berlin, Heidelberg

Summary

Many economic sectors often collect significantly less data than would be required to analyze related standard decision problems. This is because the demand for some data can be intrusive to the participants of the economy in terms of time and sensitivity. The problem of modelling and solving decision models when relevant empirical information is incomplete is addressed. First, a procedure is presented for adjusting the parameters of a model which is robust against the worst-case values of unobserved data. Second, a scenario tree approach is considered to deal with the randomness of the dynamic economic model and equilibria is computed using an interior-point algorithm. This methodology is implemented in the Australian deregulated electricity market. Although a simplified model of the market and limited information on the production side are considered, the results are very encouraging since the pattern of equilibrium prices is forecasted

Key words

Economic modelling equilibrium worst-case scenario tree interior-point methods electricity spot market 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer-Verlag Berlin Heidelberg 2007

Authors and Affiliations

  • Mercedes Esteban-Bravo
    • 1
  • Berc Rustem
    • 2
  1. 1.Department of Business AdministrationUniversidad Carlos III de MadridSpain
  2. 2.Department of ComputingImperial College LondonUK