Combined empirical mode decomposition and dynamic regression model for forecasting electricity load demand
Electricity load demand forecasting is an important element in the electric power industry for energy system planning and operation. The forecast accuracy is the main characteristic in the forecasting process. Hence, in an attempt to achieve a good forecast, combined methods of empirical mode decomp...
Saved in:
Main Author: | |
---|---|
Format: | Thesis |
Language: | English |
Published: |
2015
|
Subjects: | |
Online Access: | http://eprints.utm.my/id/eprint/53549/1/NuramirahAkromMFS2015.pdf http://eprints.utm.my/id/eprint/53549/ http://dms.library.utm.my:8080/vital/access/manager/Repository/vital:84487 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
id |
my.utm.53549 |
---|---|
record_format |
eprints |
spelling |
my.utm.535492020-07-20T01:45:07Z http://eprints.utm.my/id/eprint/53549/ Combined empirical mode decomposition and dynamic regression model for forecasting electricity load demand Akrom, Nuramirah QA Mathematics Electricity load demand forecasting is an important element in the electric power industry for energy system planning and operation. The forecast accuracy is the main characteristic in the forecasting process. Hence, in an attempt to achieve a good forecast, combined methods of empirical mode decomposition (EMD) and dynamic regression (DR), known as EMD-DR is proposed. Besides, the forecast performance of the combined model EMD and DR is compared with a single DR model. EMD is a powerful analysis technique for detecting non-stationary and nonlinear signal, while DR is a method that involves lagged external variables. The data used in this study are retrieved from half-hourly electricity demand (kW) and reactive power (var), whereby the reactive power data acts as exogenous variable for the DR method. The investigation is conducted using Statistical Analysis Software (SAS) for DR method and Matlab software for EMD. The findings reveal that the combined method, EMD-DR, give mean absolute percentage error (MAPE) 0.7237%, whereas for the DR method, 0.8074% is obtained, which suggests percentage improvement of 10.37%. 2015-03 Thesis NonPeerReviewed application/pdf en http://eprints.utm.my/id/eprint/53549/1/NuramirahAkromMFS2015.pdf Akrom, Nuramirah (2015) Combined empirical mode decomposition and dynamic regression model for forecasting electricity load demand. Masters thesis, Universiti Teknologi Malaysia, Faculty of Science. http://dms.library.utm.my:8080/vital/access/manager/Repository/vital:84487 |
institution |
Universiti Teknologi Malaysia |
building |
UTM Library |
collection |
Institutional Repository |
continent |
Asia |
country |
Malaysia |
content_provider |
Universiti Teknologi Malaysia |
content_source |
UTM Institutional Repository |
url_provider |
http://eprints.utm.my/ |
language |
English |
topic |
QA Mathematics |
spellingShingle |
QA Mathematics Akrom, Nuramirah Combined empirical mode decomposition and dynamic regression model for forecasting electricity load demand |
description |
Electricity load demand forecasting is an important element in the electric power industry for energy system planning and operation. The forecast accuracy is the main characteristic in the forecasting process. Hence, in an attempt to achieve a good forecast, combined methods of empirical mode decomposition (EMD) and dynamic regression (DR), known as EMD-DR is proposed. Besides, the forecast performance of the combined model EMD and DR is compared with a single DR model. EMD is a powerful analysis technique for detecting non-stationary and nonlinear signal, while DR is a method that involves lagged external variables. The data used in this study are retrieved from half-hourly electricity demand (kW) and reactive power (var), whereby the reactive power data acts as exogenous variable for the DR method. The investigation is conducted using Statistical Analysis Software (SAS) for DR method and Matlab software for EMD. The findings reveal that the combined method, EMD-DR, give mean absolute percentage error (MAPE) 0.7237%, whereas for the DR method, 0.8074% is obtained, which suggests percentage improvement of 10.37%. |
format |
Thesis |
author |
Akrom, Nuramirah |
author_facet |
Akrom, Nuramirah |
author_sort |
Akrom, Nuramirah |
title |
Combined empirical mode decomposition and dynamic regression model for forecasting electricity load demand |
title_short |
Combined empirical mode decomposition and dynamic regression model for forecasting electricity load demand |
title_full |
Combined empirical mode decomposition and dynamic regression model for forecasting electricity load demand |
title_fullStr |
Combined empirical mode decomposition and dynamic regression model for forecasting electricity load demand |
title_full_unstemmed |
Combined empirical mode decomposition and dynamic regression model for forecasting electricity load demand |
title_sort |
combined empirical mode decomposition and dynamic regression model for forecasting electricity load demand |
publishDate |
2015 |
url |
http://eprints.utm.my/id/eprint/53549/1/NuramirahAkromMFS2015.pdf http://eprints.utm.my/id/eprint/53549/ http://dms.library.utm.my:8080/vital/access/manager/Repository/vital:84487 |
_version_ |
1674066178939551744 |
score |
13.18916 |