Support vector machine for day ahead electricity price forecasting

Intan Azmira Binti Wan Abdul Razak, Izham Zainal Abidin, Keem Siah Yap, Titik Khawa Binti Abdul Rahman, M. Y. Lada, Anis Niza Binti Ramani, M. N.M. Nasir, Arfah Binti Ahmad

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

Abstract

Electricity price forecasting has become an important part of power system operation and planning. In a pool- based electric energy market, producers submit selling bids consisting in energy blocks and their corresponding minimum selling prices to the market operator. Meanwhile, consumers submit buying bids consisting in energy blocks and their corresponding maximum buying prices to the market operator. Hence, both producers and consumers use day ahead price forecasts to derive their respective bidding strategies to the electricity market yet reduce the cost of electricity. However, forecasting electricity prices is a complex task because price series is a non-stationary and highly volatile series. Many factors cause for price spikes such as volatility in load and fuel price as well as power import to and export from outside the market through long term contract. This paper introduces an approach of machine learning algorithm for day ahead electricity price forecasting with Least Square Support Vector Machine (LS-SVM). Previous day data of Hourly Ontario Electricity Price (HOEP), generation's price and demand from Ontario power market are used as the inputs for training data. The simulation is held using LSSVMlab in Matlab with the training and testing data of 2004. SVM that widely used for classification and regression has great generalization ability with structured risk minimization principle rather than empirical risk minimization. Moreover, same parameter settings in trained SVM give same results that absolutely reduce simulation process compared to other techniques such as neural network and time series. The mean absolute percentage error (MAPE) for the proposed model shows that SVM performs well compared to neural network.

Original languageEnglish
Title of host publicationInternational Conference on Mathematics, Engineering and Industrial Applications, ICoMEIA 2014
EditorsMohammad Fadzli Ramli, Nurshazneem Roslan, Ahmad Kadri Junoh, Maz Jamilah Masnan, Mohammad Huskhazrin Kharuddin
PublisherAmerican Institute of Physics Inc.
ISBN (Electronic)9780735413047
DOIs
Publication statusPublished - 15 May 2015
EventInternational Conference on Mathematics, Engineering and Industrial Applications, ICoMEIA 2014 - Penang, Malaysia
Duration: 28 May 201430 May 2014

Publication series

NameAIP Conference Proceedings
Volume1660
ISSN (Print)0094-243X
ISSN (Electronic)1551-7616

Other

OtherInternational Conference on Mathematics, Engineering and Industrial Applications, ICoMEIA 2014
CountryMalaysia
CityPenang
Period28/05/1430/05/14

Fingerprint

electricity
forecasting
Ontario
education
operators
machine learning
optimization
volatility
spikes
planning
energy
regression analysis
simulation
costs
causes

All Science Journal Classification (ASJC) codes

  • Ecology, Evolution, Behavior and Systematics
  • Ecology
  • Plant Science
  • Physics and Astronomy(all)
  • Nature and Landscape Conservation

Cite this

Razak, I. A. B. W. A., Zainal Abidin, I., Yap, K. S., Rahman, T. K. B. A., Lada, M. Y., Ramani, A. N. B., ... Ahmad, A. B. (2015). Support vector machine for day ahead electricity price forecasting. In M. F. Ramli, N. Roslan, A. K. Junoh, M. J. Masnan, & M. H. Kharuddin (Eds.), International Conference on Mathematics, Engineering and Industrial Applications, ICoMEIA 2014 [090021] (AIP Conference Proceedings; Vol. 1660). American Institute of Physics Inc.. https://doi.org/10.1063/1.4915865
Razak, Intan Azmira Binti Wan Abdul ; Zainal Abidin, Izham ; Yap, Keem Siah ; Rahman, Titik Khawa Binti Abdul ; Lada, M. Y. ; Ramani, Anis Niza Binti ; Nasir, M. N.M. ; Ahmad, Arfah Binti. / Support vector machine for day ahead electricity price forecasting. International Conference on Mathematics, Engineering and Industrial Applications, ICoMEIA 2014. editor / Mohammad Fadzli Ramli ; Nurshazneem Roslan ; Ahmad Kadri Junoh ; Maz Jamilah Masnan ; Mohammad Huskhazrin Kharuddin. American Institute of Physics Inc., 2015. (AIP Conference Proceedings).
@inproceedings{1148a36c3c554803b6d177388cfdff6e,
title = "Support vector machine for day ahead electricity price forecasting",
abstract = "Electricity price forecasting has become an important part of power system operation and planning. In a pool- based electric energy market, producers submit selling bids consisting in energy blocks and their corresponding minimum selling prices to the market operator. Meanwhile, consumers submit buying bids consisting in energy blocks and their corresponding maximum buying prices to the market operator. Hence, both producers and consumers use day ahead price forecasts to derive their respective bidding strategies to the electricity market yet reduce the cost of electricity. However, forecasting electricity prices is a complex task because price series is a non-stationary and highly volatile series. Many factors cause for price spikes such as volatility in load and fuel price as well as power import to and export from outside the market through long term contract. This paper introduces an approach of machine learning algorithm for day ahead electricity price forecasting with Least Square Support Vector Machine (LS-SVM). Previous day data of Hourly Ontario Electricity Price (HOEP), generation's price and demand from Ontario power market are used as the inputs for training data. The simulation is held using LSSVMlab in Matlab with the training and testing data of 2004. SVM that widely used for classification and regression has great generalization ability with structured risk minimization principle rather than empirical risk minimization. Moreover, same parameter settings in trained SVM give same results that absolutely reduce simulation process compared to other techniques such as neural network and time series. The mean absolute percentage error (MAPE) for the proposed model shows that SVM performs well compared to neural network.",
author = "Razak, {Intan Azmira Binti Wan Abdul} and {Zainal Abidin}, Izham and Yap, {Keem Siah} and Rahman, {Titik Khawa Binti Abdul} and Lada, {M. Y.} and Ramani, {Anis Niza Binti} and Nasir, {M. N.M.} and Ahmad, {Arfah Binti}",
year = "2015",
month = "5",
day = "15",
doi = "10.1063/1.4915865",
language = "English",
series = "AIP Conference Proceedings",
publisher = "American Institute of Physics Inc.",
editor = "Ramli, {Mohammad Fadzli} and Nurshazneem Roslan and Junoh, {Ahmad Kadri} and Masnan, {Maz Jamilah} and Kharuddin, {Mohammad Huskhazrin}",
booktitle = "International Conference on Mathematics, Engineering and Industrial Applications, ICoMEIA 2014",

}

Razak, IABWA, Zainal Abidin, I, Yap, KS, Rahman, TKBA, Lada, MY, Ramani, ANB, Nasir, MNM & Ahmad, AB 2015, Support vector machine for day ahead electricity price forecasting. in MF Ramli, N Roslan, AK Junoh, MJ Masnan & MH Kharuddin (eds), International Conference on Mathematics, Engineering and Industrial Applications, ICoMEIA 2014., 090021, AIP Conference Proceedings, vol. 1660, American Institute of Physics Inc., International Conference on Mathematics, Engineering and Industrial Applications, ICoMEIA 2014, Penang, Malaysia, 28/05/14. https://doi.org/10.1063/1.4915865

Support vector machine for day ahead electricity price forecasting. / Razak, Intan Azmira Binti Wan Abdul; Zainal Abidin, Izham; Yap, Keem Siah; Rahman, Titik Khawa Binti Abdul; Lada, M. Y.; Ramani, Anis Niza Binti; Nasir, M. N.M.; Ahmad, Arfah Binti.

International Conference on Mathematics, Engineering and Industrial Applications, ICoMEIA 2014. ed. / Mohammad Fadzli Ramli; Nurshazneem Roslan; Ahmad Kadri Junoh; Maz Jamilah Masnan; Mohammad Huskhazrin Kharuddin. American Institute of Physics Inc., 2015. 090021 (AIP Conference Proceedings; Vol. 1660).

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

TY - GEN

T1 - Support vector machine for day ahead electricity price forecasting

AU - Razak, Intan Azmira Binti Wan Abdul

AU - Zainal Abidin, Izham

AU - Yap, Keem Siah

AU - Rahman, Titik Khawa Binti Abdul

AU - Lada, M. Y.

AU - Ramani, Anis Niza Binti

AU - Nasir, M. N.M.

AU - Ahmad, Arfah Binti

PY - 2015/5/15

Y1 - 2015/5/15

N2 - Electricity price forecasting has become an important part of power system operation and planning. In a pool- based electric energy market, producers submit selling bids consisting in energy blocks and their corresponding minimum selling prices to the market operator. Meanwhile, consumers submit buying bids consisting in energy blocks and their corresponding maximum buying prices to the market operator. Hence, both producers and consumers use day ahead price forecasts to derive their respective bidding strategies to the electricity market yet reduce the cost of electricity. However, forecasting electricity prices is a complex task because price series is a non-stationary and highly volatile series. Many factors cause for price spikes such as volatility in load and fuel price as well as power import to and export from outside the market through long term contract. This paper introduces an approach of machine learning algorithm for day ahead electricity price forecasting with Least Square Support Vector Machine (LS-SVM). Previous day data of Hourly Ontario Electricity Price (HOEP), generation's price and demand from Ontario power market are used as the inputs for training data. The simulation is held using LSSVMlab in Matlab with the training and testing data of 2004. SVM that widely used for classification and regression has great generalization ability with structured risk minimization principle rather than empirical risk minimization. Moreover, same parameter settings in trained SVM give same results that absolutely reduce simulation process compared to other techniques such as neural network and time series. The mean absolute percentage error (MAPE) for the proposed model shows that SVM performs well compared to neural network.

AB - Electricity price forecasting has become an important part of power system operation and planning. In a pool- based electric energy market, producers submit selling bids consisting in energy blocks and their corresponding minimum selling prices to the market operator. Meanwhile, consumers submit buying bids consisting in energy blocks and their corresponding maximum buying prices to the market operator. Hence, both producers and consumers use day ahead price forecasts to derive their respective bidding strategies to the electricity market yet reduce the cost of electricity. However, forecasting electricity prices is a complex task because price series is a non-stationary and highly volatile series. Many factors cause for price spikes such as volatility in load and fuel price as well as power import to and export from outside the market through long term contract. This paper introduces an approach of machine learning algorithm for day ahead electricity price forecasting with Least Square Support Vector Machine (LS-SVM). Previous day data of Hourly Ontario Electricity Price (HOEP), generation's price and demand from Ontario power market are used as the inputs for training data. The simulation is held using LSSVMlab in Matlab with the training and testing data of 2004. SVM that widely used for classification and regression has great generalization ability with structured risk minimization principle rather than empirical risk minimization. Moreover, same parameter settings in trained SVM give same results that absolutely reduce simulation process compared to other techniques such as neural network and time series. The mean absolute percentage error (MAPE) for the proposed model shows that SVM performs well compared to neural network.

UR - http://www.scopus.com/inward/record.url?scp=85006192493&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85006192493&partnerID=8YFLogxK

U2 - 10.1063/1.4915865

DO - 10.1063/1.4915865

M3 - Conference contribution

T3 - AIP Conference Proceedings

BT - International Conference on Mathematics, Engineering and Industrial Applications, ICoMEIA 2014

A2 - Ramli, Mohammad Fadzli

A2 - Roslan, Nurshazneem

A2 - Junoh, Ahmad Kadri

A2 - Masnan, Maz Jamilah

A2 - Kharuddin, Mohammad Huskhazrin

PB - American Institute of Physics Inc.

ER -

Razak IABWA, Zainal Abidin I, Yap KS, Rahman TKBA, Lada MY, Ramani ANB et al. Support vector machine for day ahead electricity price forecasting. In Ramli MF, Roslan N, Junoh AK, Masnan MJ, Kharuddin MH, editors, International Conference on Mathematics, Engineering and Industrial Applications, ICoMEIA 2014. American Institute of Physics Inc. 2015. 090021. (AIP Conference Proceedings). https://doi.org/10.1063/1.4915865