A Class-Driven Approach Based on Long Short-Term Memory Networks for Electricity Price Scenario Generation and Reduction

Bart Stappers; N.G. Paterakis; J.K. (Koen) Kok; M. Gibescu

doi:10.1109/TPWRS.2020.2965922

A Class-Driven Approach Based on Long Short-Term Memory Networks for Electricity Price Scenario Generation and Reduction

Bart Stappers
, N.G. Paterakis (Corresponding author)
, J.K. (Koen) Kok
, M. Gibescu

Research output: Contribution to journal › Article › Academic › peer-review

35 Citations (Scopus)

407 Downloads (Pure)

Abstract

Uncertainty characterization is an essential component of decision-making problems in electricity markets. In this work, a class-driven approach is proposed to describe stochasticity. The methodology consists of a three-step process that includes a class allocation component, a generative element based on a long short-term memory neural network and an automated reduction method with a variance-based continuation criterion. The system is employed and evaluated on Dutch imbalance market prices. Test results are presented, expressing the proficiency of the approach, both in generating realistic scenario sets that reflect the erratic dynamics in the data and adequately reducing generated sets without the need to explicitly and manually predetermine the cardinality of the reduced set.

Original language	English
Article number	8957258
Pages (from-to)	3040-3050
Number of pages	11
Journal	IEEE Transactions on Power Systems
Volume	35
Issue number	4
DOIs	https://doi.org/10.1109/TPWRS.2020.2965922
Publication status	Published - Jul 2020

Funding

Manuscript received June 27, 2019; revised October 15, 2019 and December 9, 2019; accepted December 29, 2019. Date of publication January 13, 2020; date of current version June 22, 2020. This work was supported by Scholt Energy B.V., The Netherlands. Paper no. TPWRS-00919-2019. (Corresponding author: Nikolaos Paterakis.) B. Stappers is with the Department of Electrical Engineering, Eindhoven University of Technology, 5600 MB Eindhoven, The Netherlands, is also with the Scholt Energy B.V., 5555 XA Valkenswaard, The Netherlands (e-mail: [email protected]).

Funders	Funder number
Scholt Energy B.V.	TPWRS-00919-2019

Keywords

Deep learning
imbalance prices
long short-term memory (LSTM)
machine learning
recurrent neural network (RNN)
scenario generation
scenario reduction

Access to Document

10.1109/TPWRS.2020.2965922

Author versionFinal author version , 3 MB
08957258Final published version, 2.84 MBLicence: TAVERNE

Cite this

@article{a016d41a834e4bb3906bfd421a0a8b8d,

title = "A Class-Driven Approach Based on Long Short-Term Memory Networks for Electricity Price Scenario Generation and Reduction",

abstract = "Uncertainty characterization is an essential component of decision-making problems in electricity markets. In this work, a class-driven approach is proposed to describe stochasticity. The methodology consists of a three-step process that includes a class allocation component, a generative element based on a long short-term memory neural network and an automated reduction method with a variance-based continuation criterion. The system is employed and evaluated on Dutch imbalance market prices. Test results are presented, expressing the proficiency of the approach, both in generating realistic scenario sets that reflect the erratic dynamics in the data and adequately reducing generated sets without the need to explicitly and manually predetermine the cardinality of the reduced set.",

keywords = "Deep learning, imbalance prices, long short-term memory (LSTM), machine learning, recurrent neural network (RNN), scenario generation, scenario reduction",

author = "Bart Stappers and N.G. Paterakis and Kok, \{J.K. (Koen)\} and M. Gibescu",

year = "2020",

month = jul,

doi = "10.1109/TPWRS.2020.2965922",

language = "English",

volume = "35",

pages = "3040--3050",

journal = "IEEE Transactions on Power Systems",

issn = "0885-8950",

publisher = "Institute of Electrical and Electronics Engineers",

number = "4",

}

A Class-Driven Approach Based on Long Short-Term Memory Networks for Electricity Price Scenario Generation and Reduction. / Stappers, Bart; Paterakis, N.G. (Corresponding author); Kok, J.K. (Koen) et al.
In: IEEE Transactions on Power Systems, Vol. 35, No. 4, 8957258, 07.2020, p. 3040-3050.

Research output: Contribution to journal › Article › Academic › peer-review

TY - JOUR

T1 - A Class-Driven Approach Based on Long Short-Term Memory Networks for Electricity Price Scenario Generation and Reduction

AU - Stappers, Bart

AU - Paterakis, N.G.

AU - Kok, J.K. (Koen)

AU - Gibescu, M.

PY - 2020/7

Y1 - 2020/7

N2 - Uncertainty characterization is an essential component of decision-making problems in electricity markets. In this work, a class-driven approach is proposed to describe stochasticity. The methodology consists of a three-step process that includes a class allocation component, a generative element based on a long short-term memory neural network and an automated reduction method with a variance-based continuation criterion. The system is employed and evaluated on Dutch imbalance market prices. Test results are presented, expressing the proficiency of the approach, both in generating realistic scenario sets that reflect the erratic dynamics in the data and adequately reducing generated sets without the need to explicitly and manually predetermine the cardinality of the reduced set.

AB - Uncertainty characterization is an essential component of decision-making problems in electricity markets. In this work, a class-driven approach is proposed to describe stochasticity. The methodology consists of a three-step process that includes a class allocation component, a generative element based on a long short-term memory neural network and an automated reduction method with a variance-based continuation criterion. The system is employed and evaluated on Dutch imbalance market prices. Test results are presented, expressing the proficiency of the approach, both in generating realistic scenario sets that reflect the erratic dynamics in the data and adequately reducing generated sets without the need to explicitly and manually predetermine the cardinality of the reduced set.

KW - Deep learning

KW - imbalance prices

KW - long short-term memory (LSTM)

KW - machine learning

KW - recurrent neural network (RNN)

KW - scenario generation

KW - scenario reduction

UR - https://www.scopus.com/pages/publications/85086888849

U2 - 10.1109/TPWRS.2020.2965922

DO - 10.1109/TPWRS.2020.2965922

M3 - Article

AN - SCOPUS:85086888849

SN - 0885-8950

VL - 35

SP - 3040

EP - 3050

JO - IEEE Transactions on Power Systems

JF - IEEE Transactions on Power Systems

IS - 4

M1 - 8957258

ER -

A Class-Driven Approach Based on Long Short-Term Memory Networks for Electricity Price Scenario Generation and Reduction

Abstract

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this