Home // COGNITIVE 2014, The Sixth International Conference on Advanced Cognitive Technologies and Applications // View article


Creating Confidence Intervals for Reservoir Computing’s Wind Power Forecast using the Maximum likelihood method and the Distribution based method

Authors:
Breno Menezes
Mêuser Valença

Keywords: Reservoir Computing, Confidence Intervals, Maximum Likelihood Method, Distribution Fit.

Abstract:
The world is increasing the investments in electricity production from renewable sources, such as wind farms, although, the variable power production of wind farms must be balanced by other sources of energy, such as thermal units. As the amount of electric energy generated by the wind represents a higher percentage in the electric grid, it becomes more important to do accurate wind power forecasts and confidence intervals to support the system's operation and reduce its costs. In order to generate the confidence intervals, the forecasting error is often assumed to follow a Gaussian distribution. A wrong assumption can have a huge impact on the confidence intervals. This work proposes an evaluation of the forecasting error distribution generated by a Reservoir Computing network forecast in different timescales and the confidence intervals generated using the maximum likelyhood method and the distribution based method.

Pages: 172 to 177

Copyright: Copyright (c) IARIA, 2014

Publication date: May 25, 2014

Published in: conference

ISSN: 2308-4197

ISBN: 978-1-61208-340-7

Location: Venice, Italy

Dates: from May 25, 2014 to May 29, 2014