Institutional Repository
Technical University of Crete
Technical University of Crete
EN
|
EL
| URI: | http://purl.tuc.gr/dl/dias/3DA63488-0336-480C-9CC3-2EEE150543D6 | ||
| Year | 2023 | ||
| Type of Item | Diploma Work | ||
| License |
|
||
| Bibliographic Citation | Georgios Exarchakos, "Statistical analysis of wind farms’ electricity production data", Diploma Work, School of Electrical and Computer Engineering, Technical University of Crete, Chania, Greece, 2023 https://doi.org/10.26233/heallink.tuc.98373 | ||
| Appears in Collections |
In recent years, the production of electricity from renewable energy sources has intensified. This policy is helping to tackle the climate crisis. Generation systems are becoming hybrid and accurate forecasts of the loads of each plant are required to achieve proper daily energy planning. For this reason, with this thesis, power generation forecasting is done and emphasis is given to wind farms due to the unpredictable nature of wind. Initially, data are collected and then after being quality checked, the necessary time series are generated. This is followed by statistical analysis to establish their nature and detect trends, patterns and autocorrelations between them. Once the necessary transformations have been made, all possible AR, MA, ARMA, ARIMA and SARIMA models are tested and trained in order to select the most appropriate one. Based on this, out-of-sample training predictions are produced and evaluated. At the end of this process, an XGBoost machine learning model training process is developed to produce even better predictions than those provided by the statistical models. In this way we produce the required forecasts which can be used in daily energy planning.
Copyright © DIAS 2013
