Use this URL to cite or link to this record in EThOS:
Title: Statistical methods for short term wind speed prediction
Author: Mitchell, Erin
Awarding Body: Lancaster University
Current Institution: Lancaster University
Date of Award: 2013
Availability of Full Text:
Access from EThOS:
One of Earth’s most powerful natural resources is wind; harnessing its power in order to generate electricity is an ever expanding market throughout the world. In order to best utilise this resource it is important to be able to make accurate predictions of future wind speeds. This thesis focuses on accurately forecasting future wind speeds using statistical methods, in particular utilising past wind speed data and numerical weather prediction (NWP) data. Wind speed data are non-stationary time series, and the relationship between future wind speeds and NWP data may change over time. With this in mind, we look at models that can capture smooth variation in the series, specifically looking at dynamic linear models. In practice, many different forms of model can be used for predicting future wind speeds. We also look at methods for choosing between models, or for combining the predictions that different models make. We offer novel approaches that allow for abrupt changes in the choice of model based on the accuracy of recent predictions, with these approaches offering improvements in forecast accuracies. The final focus of this thesis is on developing novel methods for predicting ramp events: sudden, sharp changes in wind speeds. Predicting these events accurately is important, as ramp events can lead to substantial changes in the amount of wind energy produced. However, such events are currently not predicted well by standard forecasting models.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available