Quantile combination for the EEM Wind Power Forecasting Competition

Jethro Browell; Ciaran Gilbert; Rosemary Tawn; Leo May

doi:10.1109/EEM49802.2020.9221942

Quantile combination for the EEM Wind Power Forecasting Competition

Jethro Browell
, Ciaran Gilbert
, Rosemary Tawn
, Leo May

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

9 Citations (Scopus)

81 Downloads (Pure)

Abstract

Combining forecasts is an established strategy for improving predictions and is employed here to produce probabilistic forecasts of regional wind power production in Sweden, finishing in second place in the EEM20 Wind Power Forecasting Competition. We combine quantile forecasts from two models with different characteristics: a ‘discrete’ tree-based model and ‘smooth’ generalised additive model. Quantiles are combined via linear weighting and the resulting combination is superior than both constituent forecasts in all four regions considered.

Original language	English
Title of host publication	2020 17th International Conference on the European Energy Market (EEM)
Place of Publication	Piscataway, NJ.
Publisher	IEEE
Number of pages	6
ISBN (Print)	9781728169200
DOIs	https://doi.org/10.1109/EEM49802.2020.9221942
Publication status	Published - 13 Oct 2020
Event	17th International Conference on the European Energy Market - Duration: 16 Sept 2020 → 18 Sept 2020 https://eem20.eu/

Conference

Conference	17th International Conference on the European Energy Market
Abbreviated title	EEM20
Period	16/09/20 → 18/09/20
Internet address	https://eem20.eu/

Keywords

wind power
forecasting
Wind Power Forecasting Competition
quantile forecasts
wind energy capacity

Access to Document

10.1109/EEM49802.2020.9221942

Browell-etal-EEM-2020-Quantile-combination-for-the-EEM20-wind-power-forecasting-competitionAccepted author manuscript, 245 KB

Cite this

@inproceedings{fbcf5550ca38432c9454e35fc183f66c,

title = "Quantile combination for the EEM Wind Power Forecasting Competition",

abstract = "Combining forecasts is an established strategy for improving predictions and is employed here to produce probabilistic forecasts of regional wind power production in Sweden, finishing in second place in the EEM20 Wind Power Forecasting Competition. We combine quantile forecasts from two models with different characteristics: a {\textquoteleft}discrete{\textquoteright} tree-based model and {\textquoteleft}smooth{\textquoteright} generalised additive model. Quantiles are combined via linear weighting and the resulting combination is superior than both constituent forecasts in all four regions considered.",

keywords = "wind power, forecasting, Wind Power Forecasting Competition, quantile forecasts, wind energy capacity",

author = "Jethro Browell and Ciaran Gilbert and Rosemary Tawn and Leo May",

year = "2020",

month = oct,

day = "13",

doi = "10.1109/EEM49802.2020.9221942",

language = "English",

isbn = "9781728169200",

booktitle = "2020 17th International Conference on the European Energy Market (EEM)",

publisher = "IEEE",

note = "17th International Conference on the European Energy Market, EEM20 ; Conference date: 16-09-2020 Through 18-09-2020",

url = "https://eem20.eu/",

}

TY - GEN

T1 - Quantile combination for the EEM Wind Power Forecasting Competition

AU - Browell, Jethro

AU - Gilbert, Ciaran

AU - Tawn, Rosemary

AU - May, Leo

PY - 2020/10/13

Y1 - 2020/10/13

N2 - Combining forecasts is an established strategy for improving predictions and is employed here to produce probabilistic forecasts of regional wind power production in Sweden, finishing in second place in the EEM20 Wind Power Forecasting Competition. We combine quantile forecasts from two models with different characteristics: a ‘discrete’ tree-based model and ‘smooth’ generalised additive model. Quantiles are combined via linear weighting and the resulting combination is superior than both constituent forecasts in all four regions considered.

AB - Combining forecasts is an established strategy for improving predictions and is employed here to produce probabilistic forecasts of regional wind power production in Sweden, finishing in second place in the EEM20 Wind Power Forecasting Competition. We combine quantile forecasts from two models with different characteristics: a ‘discrete’ tree-based model and ‘smooth’ generalised additive model. Quantiles are combined via linear weighting and the resulting combination is superior than both constituent forecasts in all four regions considered.

KW - wind power

KW - forecasting

KW - Wind Power Forecasting Competition

KW - quantile forecasts

KW - wind energy capacity

UR - https://github.com/jbrowell/ProbCast

U2 - 10.1109/EEM49802.2020.9221942

DO - 10.1109/EEM49802.2020.9221942

M3 - Conference contribution book

SN - 9781728169200

BT - 2020 17th International Conference on the European Energy Market (EEM)

PB - IEEE

CY - Piscataway, NJ.

T2 - 17th International Conference on the European Energy Market

Y2 - 16 September 2020 through 18 September 2020

ER -

Quantile combination for the EEM 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

Data and Code for the EEM2020 Wind Power Forecasting Competition

2nd Place: European Energy Markets Conference 2020 Wind Power Forecasting Competition

Cite this

Quantile combination for the EEM 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

Datasets

Data and Code for the EEM2020 Wind Power Forecasting Competition

Prizes

2nd Place: European Energy Markets Conference 2020 Wind Power Forecasting Competition

Cite this