Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Prediction of Current by Artificial Neural \\ Networks in a Substation in order \\ to Schedule Thermography
KTH, School of Electrical Engineering and Computer Science (EECS), Electromagnetic Engineering. (QED-AM)ORCID iD: 0000-0002-2462-8340
KTH, School of Electrical Engineering and Computer Science (EECS), Electric Power and Energy Systems.ORCID iD: 0000-0003-4490-9278
2018 (English)In: ITISE, Granada, 2018Conference paper, Published paper (Other academic)
Abstract [en]

Thermography or infra-red imaging is a method that measures the temperature of a surface by receiving the infra-red radiation that the surface emits. Thermography is used in, for example, condition measuring of electrical equipment. It shows which parts are heated more than normally due to a higher resistance. Those parts will need maintenance.

In order to get accurate values, thermography needs a high current in the equipment. Thus it is necessary to predict when the current will be high throughout the year. Here a neural network with two layers is used for the prediction. The data set consists of the hourly currents at a point in a Swedish substation from a period of ten years.

\\

The purpose is to plan when to go to a substation to do thermography. As the prediction is done several months ahead, the outdoor temperature cannot be used. Hence only the time expressed as week, day and hour with different resolutions in the discretization, is used as an explanatory variable. With increasing resolutions in the discretization, the prediction error decreases. Adding inputs based on interaction does not improve the prediction. The results are however not satisfactory as the prediction error is large in comparison with the predicted values of the current and the prediction is biased. One reason is that the prediction should be several months ahead, so the actual temperature cannot be used.

Place, publisher, year, edition, pages
Granada, 2018.
Keywords [en]
artificial neural networks, ANN, infrared sensors, IR, prediction algorithms, substations, thermography
National Category
Probability Theory and Statistics
Research subject
Electrical Engineering
Identifiers
URN: urn:nbn:se:kth:diva-239570OAI: oai:DiVA.org:kth-239570DiVA, id: diva2:1265984
Conference
International conference on Time Series and Forecasting, September 19th-21th, 2018
Funder
SweGRIDS - Swedish Centre for Smart Grids and Energy Storage
Note

QC 20181129

Available from: 2018-11-26 Created: 2018-11-26 Last updated: 2018-11-29Bibliographically approved

Open Access in DiVA

No full text in DiVA

Authority records BETA

Westerlund, PerDimoulkas, Ilias

Search in DiVA

By author/editor
Westerlund, PerDimoulkas, Ilias
By organisation
Electromagnetic EngineeringElectric Power and Energy Systems
Probability Theory and Statistics

Search outside of DiVA

GoogleGoogle Scholar

urn-nbn

Altmetric score

urn-nbn
Total: 28 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf