Please use this identifier to cite or link to this item:
https://open.uns.ac.rs/handle/123456789/8518
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Ilić, Ljubica | en |
dc.contributor.author | Selakov, Aleksandar | en |
dc.contributor.author | Vukmirović, Goran | en |
dc.contributor.author | Erdeljan, Andrea | en |
dc.contributor.author | Kulić, Filip | en |
dc.date.accessioned | 2019-09-30T09:09:16Z | - |
dc.date.available | 2019-09-30T09:09:16Z | - |
dc.date.issued | 2013-12-01 | en |
dc.identifier.issn | 224456 | en |
dc.identifier.uri | https://open.uns.ac.rs/handle/123456789/8518 | - |
dc.description.abstract | This paper presents a novel method for short-term load forecasting (STLF), based on artificial neural network (ANN), targeted for use in large-scale systems such as distribution management system (DMS). The system comprises of a preprocessing unit (PPU) and a feed forward ANN ordered in a sequence. PPU prepares the data and feeds them as input to the ANN, which calculates the hourly load forecasts. Preprocessing of the entering data reduces the size of the input space to the ANN, which improves the generalization capability and shortens the training time of the network. Reduced dimension of the input space also diminishes the number of parameters to be set in a training procedure, allowing smaller training set, and thus online usage and adaptation. This is important for a real-world power system where a sufficient set of historical data (training points) may not always be available, for different reasons. Ease of use and fast adaptation are necessary when predictions need to carry out in a large number of nodes in the power grid. Functionality of the proposed method has been tested on recorded data from Serbian electrical utility. Results demonstrate that even with a simple configuration such as this one, fair accuracy can be achieved in forecasting the hourly load. The simplicity and reusability are very important factors for installation of the proposed system in a large-scale DMS, considering the technical requirements (e.g. training data availability, processing power and memory capacity). | en |
dc.relation.ispartof | Journal of Scientific and Industrial Research | en |
dc.title | Short-term load forecasting in large scale electrical utility using artificial neural network | en |
dc.type | Journal/Magazine Article | en |
dc.identifier.scopus | 2-s2.0-84889590172 | en |
dc.identifier.url | https://api.elsevier.com/content/abstract/scopus_id/84889590172 | en |
dc.relation.lastpage | 745 | en |
dc.relation.firstpage | 739 | en |
dc.relation.issue | 12 | en |
dc.relation.volume | 72 | en |
item.grantfulltext | none | - |
item.fulltext | No Fulltext | - |
crisitem.author.dept | Fakultet tehničkih nauka, Departman za energetiku, elektroniku i telekomunikacije | - |
crisitem.author.dept | Fakultet tehničkih nauka, Departman za računarstvo i automatiku | - |
crisitem.author.parentorg | Fakultet tehničkih nauka | - |
crisitem.author.parentorg | Fakultet tehničkih nauka | - |
Appears in Collections: | FTN Publikacije/Publications |
SCOPUSTM
Citations
9
checked on Feb 22, 2020
Page view(s)
14
Last Week
3
3
Last month
0
0
checked on May 10, 2024
Google ScholarTM
Check
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.