Very-short-term probabilistic wind power forecasts by sparse vector autoregression

Jethro Dowell, Pierre Pinson

Research output: Contribution to journalArticle

72 Citations (Scopus)

Abstract

A spatio-temporal method for producing very-short-term parametric probabilistic wind power forecasts at a large number of locations is presented. Smart grids containing tens, or hundreds, of wind generators require skilled very-short-term forecasts to operate effectively, and spatial information is highly desirable. In addition, probabilistic forecasts are widely regarded as necessary for optimal power system management as they quantify the uncertainty associated with point forecasts. Here we work within a parametric framework based on the logit-normal distribution and forecast its parameters. The location parameter for multiple wind farms is modelled as a vector-valued spatio-temporal process, and the scale parameter is tracked by modified exponential smoothing. A state-of-the-art technique for fitting sparse vector autoregressive models is employed to model the location parameter and demonstrates numerical advantages over conventional vector autoregressive models. The proposed method is tested on a dataset of 5 minute mean wind power generation at 22 wind farms in Australia. 5-minute-ahead forecasts are
produced and evaluated in terms of point and probabilistic forecast skill scores and calibration. Conventional autoregressive and vector autoregressive models serve as benchmarks.
LanguageEnglish
Pages763-770
Number of pages8
JournalIEEE Transactions on Smart Grid
Volume7
Issue number2
Early online date12 May 2015
DOIs
Publication statusPublished - 31 Mar 2016

Fingerprint

Wind power
Farms
Normal distribution
Power generation
Calibration

Keywords

  • probabilistic forecasting
  • wind power
  • power system operations
  • renewable energy

Cite this

@article{344b30a4e571471496efb1d5aa1aa707,
title = "Very-short-term probabilistic wind power forecasts by sparse vector autoregression",
abstract = "A spatio-temporal method for producing very-short-term parametric probabilistic wind power forecasts at a large number of locations is presented. Smart grids containing tens, or hundreds, of wind generators require skilled very-short-term forecasts to operate effectively, and spatial information is highly desirable. In addition, probabilistic forecasts are widely regarded as necessary for optimal power system management as they quantify the uncertainty associated with point forecasts. Here we work within a parametric framework based on the logit-normal distribution and forecast its parameters. The location parameter for multiple wind farms is modelled as a vector-valued spatio-temporal process, and the scale parameter is tracked by modified exponential smoothing. A state-of-the-art technique for fitting sparse vector autoregressive models is employed to model the location parameter and demonstrates numerical advantages over conventional vector autoregressive models. The proposed method is tested on a dataset of 5 minute mean wind power generation at 22 wind farms in Australia. 5-minute-ahead forecasts areproduced and evaluated in terms of point and probabilistic forecast skill scores and calibration. Conventional autoregressive and vector autoregressive models serve as benchmarks.",
keywords = "probabilistic forecasting, wind power, power system operations, renewable energy",
author = "Jethro Dowell and Pierre Pinson",
note = "2015 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other users, 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 components of this work in other works.",
year = "2016",
month = "3",
day = "31",
doi = "10.1109/TSG.2015.2424078",
language = "English",
volume = "7",
pages = "763--770",
journal = "IEEE Transactions on Smart Grid",
issn = "1949-3053",
number = "2",

}

Very-short-term probabilistic wind power forecasts by sparse vector autoregression. / Dowell, Jethro; Pinson, Pierre.

In: IEEE Transactions on Smart Grid, Vol. 7, No. 2, 31.03.2016, p. 763-770.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Very-short-term probabilistic wind power forecasts by sparse vector autoregression

AU - Dowell, Jethro

AU - Pinson, Pierre

N1 - 2015 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other users, 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 components of this work in other works.

PY - 2016/3/31

Y1 - 2016/3/31

N2 - A spatio-temporal method for producing very-short-term parametric probabilistic wind power forecasts at a large number of locations is presented. Smart grids containing tens, or hundreds, of wind generators require skilled very-short-term forecasts to operate effectively, and spatial information is highly desirable. In addition, probabilistic forecasts are widely regarded as necessary for optimal power system management as they quantify the uncertainty associated with point forecasts. Here we work within a parametric framework based on the logit-normal distribution and forecast its parameters. The location parameter for multiple wind farms is modelled as a vector-valued spatio-temporal process, and the scale parameter is tracked by modified exponential smoothing. A state-of-the-art technique for fitting sparse vector autoregressive models is employed to model the location parameter and demonstrates numerical advantages over conventional vector autoregressive models. The proposed method is tested on a dataset of 5 minute mean wind power generation at 22 wind farms in Australia. 5-minute-ahead forecasts areproduced and evaluated in terms of point and probabilistic forecast skill scores and calibration. Conventional autoregressive and vector autoregressive models serve as benchmarks.

AB - A spatio-temporal method for producing very-short-term parametric probabilistic wind power forecasts at a large number of locations is presented. Smart grids containing tens, or hundreds, of wind generators require skilled very-short-term forecasts to operate effectively, and spatial information is highly desirable. In addition, probabilistic forecasts are widely regarded as necessary for optimal power system management as they quantify the uncertainty associated with point forecasts. Here we work within a parametric framework based on the logit-normal distribution and forecast its parameters. The location parameter for multiple wind farms is modelled as a vector-valued spatio-temporal process, and the scale parameter is tracked by modified exponential smoothing. A state-of-the-art technique for fitting sparse vector autoregressive models is employed to model the location parameter and demonstrates numerical advantages over conventional vector autoregressive models. The proposed method is tested on a dataset of 5 minute mean wind power generation at 22 wind farms in Australia. 5-minute-ahead forecasts areproduced and evaluated in terms of point and probabilistic forecast skill scores and calibration. Conventional autoregressive and vector autoregressive models serve as benchmarks.

KW - probabilistic forecasting

KW - wind power

KW - power system operations

KW - renewable energy

UR - http://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=5165411

U2 - 10.1109/TSG.2015.2424078

DO - 10.1109/TSG.2015.2424078

M3 - Article

VL - 7

SP - 763

EP - 770

JO - IEEE Transactions on Smart Grid

T2 - IEEE Transactions on Smart Grid

JF - IEEE Transactions on Smart Grid

SN - 1949-3053

IS - 2

ER -