Quantifying the Influences on Probabilistic Wind Power Forecasts

  • 2018-08-14 15:27:22
  • Jens Schreiber, Bernhard Sick
  • 1

Abstract

In recent years, probabilistic forecasts techniques were proposed in researchas well as in applications to integrate volatile renewable energy resourcesinto the electrical grid. These techniques allow decision makers to take theuncertainty of the prediction into account and, therefore, to devise optimaldecisions, e.g., related to costs and risks in the electrical grid. However, itwas yet not studied how the input, such as numerical weather predictions,affects the model output of forecasting models in detail. Therefore, we examinethe potential influences with techniques from the field of sensitivity analysison three different black-box models to obtain insights into differences andsimilarities of these probabilistic models. The analysis shows a considerablenumber of potential influences in those models depending on, e.g., thepredicted probability and the type of model. These effects motivate the need totake various influences into account when models are tested, analyzed, orcompared. Nevertheless, results of the sensitivity analysis will allow us toselect a model with advantages in the practical application.

 

Quick Read (beta)

loading the full paper ...