Uncertainty in Long Term Predictions- A New Approach to Reduction
Author
Siddabathini, Praveen
Term
10. term
Education
Publication year
2008
Pages
77
Abstract
Langsigtet estimering af vindressourcer viser ofte afvigelser i vindhastighed, hvilket skaber usikkerhed. Denne usikkerhed påvirker beregninger af energiproduktion og dermed en vindparks økonomiske bæredygtighed og øger investeringsrisikoen. Studiet undersøger, om usikkerhed fra langsigtede forudsigelser kan reduceres. Først opstilles en indledende proces, hvor flere cases analyseres. Analysen viser, at usikkerheden er knyttet til uensartede tidsperioder, der bruges i forudsigelserne. På den baggrund introduceres en konsistent periode-metode, hvor forudsigelser baseres på ensartede, sammenlignelige perioder i måledata. Når metoden anvendes, reduceres usikkerheden i de analyserede cases. Konklusionen er, at tilgangen kan reducere usikkerhed i langsigtede forudsigelser ved at bruge korttidsmålte tidsserier (løbende målinger over tid).
Long-term estimates of wind resources often show deviations in wind speed, creating uncertainty. This uncertainty affects estimates of energy production and the economic viability of wind farms, increasing investment risk. The study asks whether uncertainty from long-term predictions can be reduced. It sets up an initial process and analyzes several cases. The analysis finds that uncertainty is linked to using inconsistent time periods in predictions. Based on this, the study introduces a consistent period method, which makes predictions using uniform, comparable periods in the measurement data. When this method is applied, uncertainty is reduced in the case studies. The conclusion is that this approach can reduce uncertainty in long-term predictions by using short-term measured time series (continuous measurements over time).
[This abstract was generated with the help of AI]
Keywords
Documents