magistrsko delo
Urban Kos (Author), Sašo Karakatič (Mentor)

Abstract

Predvidevanje porabe električne energije predstavlja zelo pomemben člen v elektroenergetski industriji, saj lahko pripomore k optimizaciji proizvodnje. S pomočjo strojnega učenja, natančneje rekurentnih nevronskih mrež, je mogoče natančno napovedati električno energijo. Veliko vlogo pri napovedovanju igrajo kakovost in količina podatkov ter arhitektura in nastavitve nevronske mreže. V teoretičnem delu je podrobno opisana nevronska mreža in njeni osnovni gradniki, kjer je bilo največ pozornosti posvečene rekurentnim mrežam, praktični del pa prikazuje izvedbo eksperimenta napovedovanja porabe električne energije z rekurentnimi nevronskimi mrežami z različno arhitekturo in podatki.

Keywords

rekurentne nevronske mreže;električna energija;napovedovanje električne energije;magistrske naloge;

Data

Language: Slovenian
Year of publishing:
Typology: 2.09 - Master's Thesis
Organization: UM FERI - Faculty of Electrical Engineering and Computer Science
Publisher: [U. Kos]
UDC: [004.832:519.216]:621.31(043.2)
COBISS: 27115523 Link will open in a new window
Views: 583
Downloads: 102
Average score: 0 (0 votes)
Metadata: JSON JSON-RDF JSON-LD TURTLE N-TRIPLES XML RDFA MICRODATA DC-XML DC-RDF RDF

Other data

Secondary language: English
Secondary title: Predicting power consumption with recurrent neural networks
Secondary abstract: The anticipation of electricity consumption represents a very important link in the electrical energy industry as it can help optimize production. With the help of machine learning, more precisely recurrent neural networks, electricity can be accurately predicted. The quality and quantity of data, as well as the architecture and settings of the neural network play a big role in forecasting. The theoretical part describes in detail the neural network and its basic building blocks, where the greatest attention was paid to the recurrent parts of the network and the practical part shows the implementation of an experiment for the prediction of electricity consumption with recurrent neural networks with different architecture and data.
Secondary keywords: recurrent neural networks;electricity;electricity forecasting;
Type (COBISS): Master's thesis/paper
Thesis comment: Univ. v Mariboru, Fak. za elektrotehniko, računalništvo in informatiko, Informatika in tehnologije komuniciranja
Pages: XII, 124 str.
ID: 11567992