Journal metrics

Journal metrics

  • CiteScore value: 1.02 CiteScore 1.02
  • SNIP value: 0.614 SNIP 0.614
  • SJR value: 0.435 SJR 0.435
  • IPP value: 0.97 IPP 0.97
  • h5-index value: 11 h5-index 11
  • Scimago H index value: 32 Scimago H index 32
Volume 45 | Copyright
Adv. Geosci., 45, 13-17, 2018
https://doi.org/10.5194/adgeo-45-13-2018
© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.

  19 Jul 2018

19 Jul 2018

Probabilistic short term wind power forecasts using deep neural networks with discrete target classes

Martin Felder1, Frank Sehnke1, Kay Ohnmeiß1, Leon Schröder1, Constantin Junk2, and Anton Kaifel1 Martin Felder et al.
  • 1Zentrum für Sonnenenergie- und Wasserstoff-Forschung Baden-Württemberg (ZSW), Meitnerstraße 1, 70563 Stuttgart, Germany
  • 2UBIMET GmbH, Schönfeldstraße 8, 76131 Karlsruhe, Germany

Abstract. Usually, neural networks trained on historical feed-in time series of wind turbines deterministically predict power output over the next hours to days. Here, the training goal is to minimise a scalar cost function, often the root mean square error (RMSE) between network output and target values. Yet similar to the analog ensemble (AnEn) method, the training algorithm can also be adapted to analyse the uncertainty of the power output from the spread of possible targets found in the historical data for a certain meteorological situation. In this study, the uncertainty estimate is achieved by discretising the continuous time series of power targets into several bins (classes). For each forecast horizon, a neural network then predicts the probability of power output falling into each of the bins, resulting in an empirical probability distribution. Similiar to the AnEn method, the proposed method avoids the use of costly numerical weather prediction (NWP) ensemble runs, although a selection of several deterministic NWP forecasts as input is helpful. Using state-of-the-art deep learning technology, we applied our method to a large region and a single wind farm. MAE scores of the 50-percentile were on par with or better than comparable deterministic forecasts. The corresponding Continuous Ranked Probability Score (CRPS) was even lower. Future work will investigate the overdispersiveness sometimes observed, and extend the method to solar power forecasts.

Download & links
Download
Short summary
In current applications of short term wind power forecasts, reliability information for the forecast is important because it helps decision making. Forecast errors are often estimated using ensembles of slightly disturbed weather predictions. Since this is very computation intensive, we propose a new method: A Deep Neural Network learns how reliable to forecast is, by comparing the current weather forecast with historical forecasts and the corresponding wind power production.
In current applications of short term wind power forecasts, reliability information for the...
Citation
Share