Development of a Computational Intelligent System for Short-Term Electric Load Forecasting

Simple item page
dc.contributor.authorFaleye, Abimbola Rashidat
dc.coverage.geographicalNigeriaen_US
dc.date.accessioned2015-09-29T11:01:23Z
dc.date.accessioned2018-10-27T12:25:09Z
dc.date.available2015-09-29T11:01:23Z
dc.date.available2018-10-27T12:25:09Z
dc.date.issued2015-09-29
dc.degree.awardMaster of Science in Computer Scienceen_US
dc.departmentComputer Scienceen_US
dc.description.abstractThis study elicited information related to Short-Term Electric Load Forecasting (STELF) and developed an intelligent system based on the information. This was with a view to setting the basis for implementing a commercial software for electric load forecasting. Historical data on Short-Term Electric Load Forecasting for three years ranging from 2004 to 2006 (being the available data) and information on the process involved in Short-Term Electric Load Forecasting were collected from the National Control Centre, Osogbo using interview, observation and contextual inquiries as well as user diary (system logbook). The knowledge embedded in the data and information collected were elicited and represented using fuzzy logic based rules. The fuzzy knowledge space was developed using fuzzy logic tool in Matlab 7 and Fuzzy Decision Tree software (FID 3.4). The data for the year 2004 and 2005 was used to develop the system. In order to evaluate the performance of the proposed model, electric load forecasting was performed on National Control Centre data. The randomly selected data from 2004 and 2005 was used for model validation while the data for 2006 was used for model verification. An intelligent system was developed and used to produce a 24-hour ahead forecasting of electric load. The prediction results from the proposed model (Fload) and that of the conventional model (F) were compared with the actual load based on fractional errors computed. The fractional errors were the variations from the actual load. The average fractional errors for Fload and F for the different periods are: January 1st 2004 (0.12 and 0.45), June 5th 2004 (24.91 and 42.95), March 2nd 2005 (0.14 and 1.10), September 4th 2005 (0.22 and 0.88) and January Ist 2006 (0.22 and 0.78). The average fractional forecast errors for the proposed model were less than that of the conventional model to validate the effectiveness of the proposed approach. The study concluded that the use of fuzzy logic in short-term electric load forecasting gave more accurate results than using the conventional model, hence the advantage of the proposed model over the conventional model.en_US
dc.facultiesTechnologyen_US
dc.format.filetypePDFen_US
dc.identifier.urihttp://localhost:8080/xmlui/handle/123456789/2134
dc.language.isoenen_US
dc.subjectShort-term electric load forecastingen_US
dc.subjectElectric load forecastingen_US
dc.titleDevelopment of a Computational Intelligent System for Short-Term Electric Load Forecastingen_US
dc.typeThesisen_US
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
Development of a Computational Intelligent System by Faleye.pdf
Size:
32.99 KB
Format:
Adobe Portable Document Format
Description:
Download
License bundle
Now showing 1 - 2 of 2
Thumbnail Image
Name:
license.txt
Size:
1.71 KB
Format:
Item-specific license agreed upon to submission
Description:
Download
Thumbnail Image
Name:
license.txt
Size:
0 B
Format:
Item-specific license agreed upon to submission
Description:
Download
Collections
Theses and Dissertations