Malaysian peak daily load forecasting

Fadhilah Abd Razak, Hashim H. Amir, Izham Zainal Abidin, Shitan Mahendran

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Time series analysis has been applied intensively and sophisticatedly to model and forecast many problems in the biological, physical and environmental phenomena of interest. This fact accounts for the basic engineering problem in forecasting the daily peak system load to use time series analysis. ARMA and Regression with ARMA errors models are among the times series models considered. ANFIS, a hybrid model from neural network is also discussed as for comparison purposes. The main interest of the forecasts consists of three days up to seven days ahead predictions for daily data. The objective is to find an appropriate model for forecasting the Malaysian peak daily demand of electricity. The pure autoregressive model with an order 2 or AR (2) has the minimum AIC statistic value compared with other ARMA models. AR (2) model recorded the value for the mean absolute percentage error (MAPE) as 1.27 % for the prediction of 3 days ahead from Jan 1 to 3 , 2005. Besides AR(2) model, Regression model with ARMA errors and ANFIS were found to be among the best forecasting models for weekdays with MAPE value from 0.1% to 3%.

Original languageEnglish
Title of host publicationSCOReD2009 - Proceedings of 2009 IEEE Student Conference on Research and Development
Pages392-394
Number of pages3
DOIs
Publication statusPublished - 2009
Event2009 IEEE Student Conference on Research and Development, SCOReD2009 - Serdang, Malaysia
Duration: 16 Nov 200918 Nov 2009

Other

Other2009 IEEE Student Conference on Research and Development, SCOReD2009
CountryMalaysia
CitySerdang
Period16/11/0918/11/09

Fingerprint

Time series analysis
Time series
Electricity
Statistics
Neural networks

All Science Journal Classification (ASJC) codes

  • Biomedical Engineering
  • Control and Systems Engineering
  • Electrical and Electronic Engineering

Cite this

Abd Razak, F., Amir, H. H., Zainal Abidin, I., & Mahendran, S. (2009). Malaysian peak daily load forecasting. In SCOReD2009 - Proceedings of 2009 IEEE Student Conference on Research and Development (pp. 392-394). [5442993] https://doi.org/10.1109/SCORED.2009.5442993
Abd Razak, Fadhilah ; Amir, Hashim H. ; Zainal Abidin, Izham ; Mahendran, Shitan. / Malaysian peak daily load forecasting. SCOReD2009 - Proceedings of 2009 IEEE Student Conference on Research and Development. 2009. pp. 392-394
@inproceedings{8aa030a6656e4d6a8ab7e78ef55de2bf,
title = "Malaysian peak daily load forecasting",
abstract = "Time series analysis has been applied intensively and sophisticatedly to model and forecast many problems in the biological, physical and environmental phenomena of interest. This fact accounts for the basic engineering problem in forecasting the daily peak system load to use time series analysis. ARMA and Regression with ARMA errors models are among the times series models considered. ANFIS, a hybrid model from neural network is also discussed as for comparison purposes. The main interest of the forecasts consists of three days up to seven days ahead predictions for daily data. The objective is to find an appropriate model for forecasting the Malaysian peak daily demand of electricity. The pure autoregressive model with an order 2 or AR (2) has the minimum AIC statistic value compared with other ARMA models. AR (2) model recorded the value for the mean absolute percentage error (MAPE) as 1.27 {\%} for the prediction of 3 days ahead from Jan 1 to 3 , 2005. Besides AR(2) model, Regression model with ARMA errors and ANFIS were found to be among the best forecasting models for weekdays with MAPE value from 0.1{\%} to 3{\%}.",
author = "{Abd Razak}, Fadhilah and Amir, {Hashim H.} and {Zainal Abidin}, Izham and Shitan Mahendran",
year = "2009",
doi = "10.1109/SCORED.2009.5442993",
language = "English",
isbn = "9781424451876",
pages = "392--394",
booktitle = "SCOReD2009 - Proceedings of 2009 IEEE Student Conference on Research and Development",

}

Abd Razak, F, Amir, HH, Zainal Abidin, I & Mahendran, S 2009, Malaysian peak daily load forecasting. in SCOReD2009 - Proceedings of 2009 IEEE Student Conference on Research and Development., 5442993, pp. 392-394, 2009 IEEE Student Conference on Research and Development, SCOReD2009, Serdang, Malaysia, 16/11/09. https://doi.org/10.1109/SCORED.2009.5442993

Malaysian peak daily load forecasting. / Abd Razak, Fadhilah; Amir, Hashim H.; Zainal Abidin, Izham; Mahendran, Shitan.

SCOReD2009 - Proceedings of 2009 IEEE Student Conference on Research and Development. 2009. p. 392-394 5442993.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Malaysian peak daily load forecasting

AU - Abd Razak, Fadhilah

AU - Amir, Hashim H.

AU - Zainal Abidin, Izham

AU - Mahendran, Shitan

PY - 2009

Y1 - 2009

N2 - Time series analysis has been applied intensively and sophisticatedly to model and forecast many problems in the biological, physical and environmental phenomena of interest. This fact accounts for the basic engineering problem in forecasting the daily peak system load to use time series analysis. ARMA and Regression with ARMA errors models are among the times series models considered. ANFIS, a hybrid model from neural network is also discussed as for comparison purposes. The main interest of the forecasts consists of three days up to seven days ahead predictions for daily data. The objective is to find an appropriate model for forecasting the Malaysian peak daily demand of electricity. The pure autoregressive model with an order 2 or AR (2) has the minimum AIC statistic value compared with other ARMA models. AR (2) model recorded the value for the mean absolute percentage error (MAPE) as 1.27 % for the prediction of 3 days ahead from Jan 1 to 3 , 2005. Besides AR(2) model, Regression model with ARMA errors and ANFIS were found to be among the best forecasting models for weekdays with MAPE value from 0.1% to 3%.

AB - Time series analysis has been applied intensively and sophisticatedly to model and forecast many problems in the biological, physical and environmental phenomena of interest. This fact accounts for the basic engineering problem in forecasting the daily peak system load to use time series analysis. ARMA and Regression with ARMA errors models are among the times series models considered. ANFIS, a hybrid model from neural network is also discussed as for comparison purposes. The main interest of the forecasts consists of three days up to seven days ahead predictions for daily data. The objective is to find an appropriate model for forecasting the Malaysian peak daily demand of electricity. The pure autoregressive model with an order 2 or AR (2) has the minimum AIC statistic value compared with other ARMA models. AR (2) model recorded the value for the mean absolute percentage error (MAPE) as 1.27 % for the prediction of 3 days ahead from Jan 1 to 3 , 2005. Besides AR(2) model, Regression model with ARMA errors and ANFIS were found to be among the best forecasting models for weekdays with MAPE value from 0.1% to 3%.

UR - http://www.scopus.com/inward/record.url?scp=77952665108&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77952665108&partnerID=8YFLogxK

U2 - 10.1109/SCORED.2009.5442993

DO - 10.1109/SCORED.2009.5442993

M3 - Conference contribution

SN - 9781424451876

SP - 392

EP - 394

BT - SCOReD2009 - Proceedings of 2009 IEEE Student Conference on Research and Development

ER -

Abd Razak F, Amir HH, Zainal Abidin I, Mahendran S. Malaysian peak daily load forecasting. In SCOReD2009 - Proceedings of 2009 IEEE Student Conference on Research and Development. 2009. p. 392-394. 5442993 https://doi.org/10.1109/SCORED.2009.5442993