Probabilistic Wind Power Forecasting Model - Case Study Datasets

<p>Two datasets are provided. </p><p>The file named “<i>InputData.xlsx</i>’ is related to the wind power data used to train the model. The data consisted of half-hourly aggregated wind power values from wind farms across the Great Britain for the period of 01/03/2014–30/9/2014. Column headings are included to explain the variables.</p><p>The file named “Outputs.mat’ is related to model outputs. This file can be opened only if you have installed MATLAB at your system. </p><p>Once opened, a file of type “struct” includes all the forecast for different cases. This file is divided into fields, each of them representing a different scenario. So, the fields are divided based on the:</p><p>Number of Generated Forecasts;</p><p>Number of Magnitude Classes used for the training process;<br></p><p>The update frequency of the model (forecasting process)<br></p><p></p><p>For example the file “<i>Outputs.Num_Of_Forecasts1000.MagnitudeClasses10.UpdateFreq16</i>” shows the generated 1000 weekly forecasts, when using 10 Magnitude classes for the training and updating the model every 16 half hours.</p><p>Column headings are provided to help understand its column. The actual data are provided in order to compare the generated forecasts with them and evaluate the performance of the model.</p><p>Research results based upon these data are published at  </p><p> </p>