Cluster-based regime-switching AR for the EEM 2017 Wind Power Forecasting Competition

Jethro Browell; Ciaran P Gilbert

Cluster-based regime-switching AR for the EEM 2017 Wind Power Forecasting Competition

Jethro Browell, Ciaran P Gilbert

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution book

217 Downloads (Pure)

Abstract

This paper describes the regime-switching autoregressive models used to win the EEM 2017 Wind Power Forecasting Competition. The competition required participants to produce daily forecast wind power production for a portfolio of wind farms from 2 to 38 hours-ahead based on historic generation and numerical weather prediction analysis data only. The regimes used in the methodology presented are defined on the previous day’s weather conditions using the k-medians clustering algorithm. Cross-validation is used to identify models with the best predictive power from a pool of candidate models. The final methodology produced a final weighted mean absolute error 4.5% lower than the second place team during the two-week competition period.

Original language	English
Title of host publication	2017 14th International Conference on the European Electricity Market Conference (EEM)
Place of Publication	Piscataway
Publisher	IEEE
ISBN (Print)	9781509054992
Publication status	Accepted/In press - 19 May 2017
Event	14th International Conference on the European Energy Market - Technische Universität Dresden, Dresden, Germany Duration: 6 Jun 2017 → 9 Jun 2017

Conference

Conference	14th International Conference on the European Energy Market
Abbreviated title	EEM 17
Country/Territory	Germany
City	Dresden
Period	6/06/17 → 9/06/17

Keywords

wind power
forecasting
time series
clustering
autoregression
regime switching

Access to Document

Browell-Gilbert-EEM-2017-Cluster-based-regime-switching-AR-for-the-EEM-2017-Wind-Power-Forecasting-CompetitionAccepted author manuscript, 574 KB

Cite this

@inproceedings{9937e1a136234d51b61b792d7f4c70b4,

title = "Cluster-based regime-switching AR for the EEM 2017 Wind Power Forecasting Competition",

abstract = "This paper describes the regime-switching autoregressive models used to win the EEM 2017 Wind Power Forecasting Competition. The competition required participants to produce daily forecast wind power production for a portfolio of wind farms from 2 to 38 hours-ahead based on historic generation and numerical weather prediction analysis data only. The regimes used in the methodology presented are defined on the previous day{\textquoteright}s weather conditions using the k-medians clustering algorithm. Cross-validation is used to identify models with the best predictive power from a pool of candidate models. The final methodology produced a final weighted mean absolute error 4.5% lower than the second place team during the two-week competition period.",

keywords = "wind power, forecasting, time series, clustering, autoregression, regime switching",

author = "Jethro Browell and Gilbert, {Ciaran P}",

note = "{\textcopyright} 2017 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.; 14th International Conference on the European Energy Market, EEM 17 ; Conference date: 06-06-2017 Through 09-06-2017",

year = "2017",

month = may,

day = "19",

language = "English",

isbn = "9781509054992",

booktitle = "2017 14th International Conference on the European Electricity Market Conference (EEM)",

publisher = "IEEE",

}

TY - GEN

T1 - Cluster-based regime-switching AR for the EEM 2017 Wind Power Forecasting Competition

AU - Browell, Jethro

AU - Gilbert, Ciaran P

N1 - © 2017 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

PY - 2017/5/19

Y1 - 2017/5/19

N2 - This paper describes the regime-switching autoregressive models used to win the EEM 2017 Wind Power Forecasting Competition. The competition required participants to produce daily forecast wind power production for a portfolio of wind farms from 2 to 38 hours-ahead based on historic generation and numerical weather prediction analysis data only. The regimes used in the methodology presented are defined on the previous day’s weather conditions using the k-medians clustering algorithm. Cross-validation is used to identify models with the best predictive power from a pool of candidate models. The final methodology produced a final weighted mean absolute error 4.5% lower than the second place team during the two-week competition period.

AB - This paper describes the regime-switching autoregressive models used to win the EEM 2017 Wind Power Forecasting Competition. The competition required participants to produce daily forecast wind power production for a portfolio of wind farms from 2 to 38 hours-ahead based on historic generation and numerical weather prediction analysis data only. The regimes used in the methodology presented are defined on the previous day’s weather conditions using the k-medians clustering algorithm. Cross-validation is used to identify models with the best predictive power from a pool of candidate models. The final methodology produced a final weighted mean absolute error 4.5% lower than the second place team during the two-week competition period.

KW - wind power

KW - forecasting

KW - time series

KW - clustering

KW - autoregression

KW - regime switching

UR - http://eem2017.com/

UR - http://ieeexplore.ieee.org/servlet/opac?punumber=1002121

M3 - Conference contribution book

SN - 9781509054992

BT - 2017 14th International Conference on the European Electricity Market Conference (EEM)

PB - IEEE

CY - Piscataway

T2 - 14th International Conference on the European Energy Market

Y2 - 6 June 2017 through 9 June 2017

ER -

Cluster-based regime-switching AR for the EEM 2017 Wind Power Forecasting Competition

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

EPSRC Centre for Doctoral Training in Wind & Marine Energy Systems | Gilbert, Ciaran

Doctoral Training Partnership (DTP - University of Strathclyde)

EEM 2017 Wind Power Forecasting Competition: Competition Data and Code used by the Winning Team

1st Place: European Energy Markets Conference 2017 Wind Power Forecasting Competition

EPSRC Doctoral Prize

Cite this

Cluster-based regime-switching AR for the EEM 2017 Wind Power Forecasting Competition

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Projects

EPSRC Centre for Doctoral Training in Wind & Marine Energy Systems | Gilbert, Ciaran

Doctoral Training Partnership (DTP - University of Strathclyde)

Datasets

EEM 2017 Wind Power Forecasting Competition: Competition Data and Code used by the Winning Team

Prizes

1st Place: European Energy Markets Conference 2017 Wind Power Forecasting Competition

EPSRC Doctoral Prize

Cite this