Χωροχρονική μοντελοποίηση του σφάλματος της πρόβλεψης της αιολικής ισχύος με χρήση νευρωνικών δικτύων

Σούλης, Ιωάννης; Soulis, Ioannis

dc.contributor.author	Σούλης, Ιωάννης	el
dc.contributor.author	Soulis, Ioannis	en
dc.date.accessioned	2015-09-09T08:32:34Z
dc.date.available	2015-09-09T08:32:34Z
dc.date.issued	2015-09-09
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/41227
dc.identifier.uri	http://dx.doi.org/10.26240/heal.ntua.9885
dc.rights	Default License
dc.subject	Αιολικές προβλέψεις	el
dc.subject	Wind power predictions	en
dc.subject	Νευρωνικά δίκτυα	el
dc.subject	Μοντελοποίηση σφάλματος	el
dc.subject	Multilayer perceptrons	en
dc.subject	Error modeling	en
dc.title	Χωροχρονική μοντελοποίηση του σφάλματος της πρόβλεψης της αιολικής ισχύος με χρήση νευρωνικών δικτύων	el
heal.type	bachelorThesis
heal.classification	Ανανεώσιμες πηγές ενέργειας	el
heal.language	el
heal.access	free
heal.recordProvider	ntua	el
heal.publicationDate	2015-07-13
heal.abstract	The growing participation of Renewable Energy Sources in energy market causes necessities for precise and on-time foreknowledge of the underlying produced power. In particular, due to the great variability of wind features, the need for optimal wind power integration led to development of state-of-the-art forecasting techniques that transform historical data and numerical weather predictions to power forecasts. In this context, this diploma thesis looks into the most recent methods and focuses in studying the spatial-temporal propagation of errors of spot forecasts for a wind farm or a region. Under the assumption that errors travel from a location to another according to the wind direction, it is possible to estimate future errors and, thus, improve a base forecasting method. The thesis proposes Multilayer Perceptrons as the Machine Learning technique that, when trained properly, can correlate efficiently the propagated errors. The proposed techinque is tested at first at Crete’s wind farms and then applied in the cases of Peloponnese, Euvoia and Rhodes.	en
heal.advisorName	Χατζηαργυρίου, Νικόλαος	el
heal.committeeMemberName	Παπαθανασίου, Σταύρος	el
heal.committeeMemberName	Γεωργιλάκης, Παύλος	el
heal.academicPublisher	Εθνικό Μετσόβιο Πολυτεχνείο. Σχολή Ηλεκτρολόγων Μηχανικών και Μηχανικών Υπολογιστών. Τομέας Ηλεκτρικής Ισχύος	el
heal.academicPublisherID	ntua
heal.numberOfPages	109 σ.
heal.fullTextAvailability	true