Day Ahead Market price scenario generation using a Combined Quantile Regression Deep Neural Network and a Non-parametric Bayesian Network: A framework for risk-based Demand Response

Ties van der Heijden; Peter Palensky; Nick van de Giesen; Edo Abraham

doi:10.1109/POWERCON53406.2022.9929940

Day Ahead Market price scenario generation using a Combined Quantile Regression Deep Neural Network and a Non-parametric Bayesian Network: A framework for risk-based Demand Response

Ties van der Heijden, Peter Palensky, Nick van de Giesen, Edo Abraham

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

2 Citations (Scopus)

57 Downloads (Pure)

Abstract

In this manuscript we propose a methodology to generate electricity price scenarios from probabilistic forecasts. Using a Combined Quantile Regression Deep Neural Network, we forecast hourly marginal price distribution quantiles for the DAM on which we fit parametric distributions. A Non-parametric Bayesian Network (BN) is applied to sample from these distributions while using the observed rank-correlation in the data to condition the samples. This results in a methodology that can create an unbounded amount of price-scenarios that obey both the forecast hourly marginal price distributions and the observed dependencies between the hourly prices in the data. The BN makes no assumptions on the marginal distribution, allowing us to flexibly change the marginal distributions of hourly forecasts while maintaining the dependency structure.

Original language	English
Title of host publication	Proceedings of the 2022 IEEE International Conference on Power Systems Technology (POWERCON)
Publisher	IEEE
Pages	1-5
Number of pages	5
ISBN (Electronic)	978-1-6654-1775-4
ISBN (Print)	978-1-6654-1776-1
DOIs	https://doi.org/10.1109/POWERCON53406.2022.9929940
Publication status	Published - 2022
Event	2022 IEEE International Conference on Power Systems Technology (POWERCON) - Kuala Lumpur, Malaysia Duration: 12 Sept 2022 → 14 Sept 2022

Conference

Conference	2022 IEEE International Conference on Power Systems Technology (POWERCON)
Country/Territory	Malaysia
City	Kuala Lumpur
Period	12/09/22 → 14/09/22

Bibliographical note

Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care

Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.

Keywords

Probabilistic electricity price forecasting
scenario generation
deep neural network
non-parametric bayesian networks
quantile regression
probabilistic forecasting
day ahead market
demand response

Access to Document

10.1109/POWERCON53406.2022.9929940

Day_Ahead_Market_price_scenario_generation_using_a_Combined_Quantile_Regression_Deep_Neural_Network_and_a_Non-parametric_Bayesian_NetworkFinal published version, 525 KB

Cite this

van der Heijden, T., Palensky, P., van de Giesen, N., & Abraham, E. (2022). Day Ahead Market price scenario generation using a Combined Quantile Regression Deep Neural Network and a Non-parametric Bayesian Network: A framework for risk-based Demand Response. In Proceedings of the 2022 IEEE International Conference on Power Systems Technology (POWERCON) (pp. 1-5). IEEE. https://doi.org/10.1109/POWERCON53406.2022.9929940

van der Heijden, Ties ; Palensky, Peter ; van de Giesen, Nick et al. / Day Ahead Market price scenario generation using a Combined Quantile Regression Deep Neural Network and a Non-parametric Bayesian Network : A framework for risk-based Demand Response. Proceedings of the 2022 IEEE International Conference on Power Systems Technology (POWERCON). IEEE, 2022. pp. 1-5

@inproceedings{06a286c4f46f400d9185e6b8412a95b6,

title = "Day Ahead Market price scenario generation using a Combined Quantile Regression Deep Neural Network and a Non-parametric Bayesian Network: A framework for risk-based Demand Response",

abstract = "In this manuscript we propose a methodology to generate electricity price scenarios from probabilistic forecasts. Using a Combined Quantile Regression Deep Neural Network, we forecast hourly marginal price distribution quantiles for the DAM on which we fit parametric distributions. A Non-parametric Bayesian Network (BN) is applied to sample from these distributions while using the observed rank-correlation in the data to condition the samples. This results in a methodology that can create an unbounded amount of price-scenarios that obey both the forecast hourly marginal price distributions and the observed dependencies between the hourly prices in the data. The BN makes no assumptions on the marginal distribution, allowing us to flexibly change the marginal distributions of hourly forecasts while maintaining the dependency structure.",

keywords = "Probabilistic electricity price forecasting, scenario generation, deep neural network, non-parametric bayesian networks, quantile regression, probabilistic forecasting, day ahead market, demand response",

author = "{van der Heijden}, Ties and Peter Palensky and {van de Giesen}, Nick and Edo Abraham",

note = "Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.; 2022 IEEE International Conference on Power Systems Technology (POWERCON) ; Conference date: 12-09-2022 Through 14-09-2022",

year = "2022",

doi = "10.1109/POWERCON53406.2022.9929940",

language = "English",

isbn = "978-1-6654-1776-1",

pages = "1--5",

booktitle = "Proceedings of the 2022 IEEE International Conference on Power Systems Technology (POWERCON)",

publisher = "IEEE",

address = "United States",

}

van der Heijden, T , Palensky, P , van de Giesen, N & Abraham, E 2022, Day Ahead Market price scenario generation using a Combined Quantile Regression Deep Neural Network and a Non-parametric Bayesian Network: A framework for risk-based Demand Response. in Proceedings of the 2022 IEEE International Conference on Power Systems Technology (POWERCON). IEEE, pp. 1-5, 2022 IEEE International Conference on Power Systems Technology (POWERCON), Kuala Lumpur, Malaysia, 12/09/22. https://doi.org/10.1109/POWERCON53406.2022.9929940

Day Ahead Market price scenario generation using a Combined Quantile Regression Deep Neural Network and a Non-parametric Bayesian Network: A framework for risk-based Demand Response. / van der Heijden, Ties ; Palensky, Peter ; van de Giesen, Nick et al.
Proceedings of the 2022 IEEE International Conference on Power Systems Technology (POWERCON). IEEE, 2022. p. 1-5.

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

TY - GEN

T1 - Day Ahead Market price scenario generation using a Combined Quantile Regression Deep Neural Network and a Non-parametric Bayesian Network

T2 - 2022 IEEE International Conference on Power Systems Technology (POWERCON)

AU - van der Heijden, Ties

AU - Palensky, Peter

AU - van de Giesen, Nick

AU - Abraham, Edo

N1 - Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.

PY - 2022

Y1 - 2022

N2 - In this manuscript we propose a methodology to generate electricity price scenarios from probabilistic forecasts. Using a Combined Quantile Regression Deep Neural Network, we forecast hourly marginal price distribution quantiles for the DAM on which we fit parametric distributions. A Non-parametric Bayesian Network (BN) is applied to sample from these distributions while using the observed rank-correlation in the data to condition the samples. This results in a methodology that can create an unbounded amount of price-scenarios that obey both the forecast hourly marginal price distributions and the observed dependencies between the hourly prices in the data. The BN makes no assumptions on the marginal distribution, allowing us to flexibly change the marginal distributions of hourly forecasts while maintaining the dependency structure.

AB - In this manuscript we propose a methodology to generate electricity price scenarios from probabilistic forecasts. Using a Combined Quantile Regression Deep Neural Network, we forecast hourly marginal price distribution quantiles for the DAM on which we fit parametric distributions. A Non-parametric Bayesian Network (BN) is applied to sample from these distributions while using the observed rank-correlation in the data to condition the samples. This results in a methodology that can create an unbounded amount of price-scenarios that obey both the forecast hourly marginal price distributions and the observed dependencies between the hourly prices in the data. The BN makes no assumptions on the marginal distribution, allowing us to flexibly change the marginal distributions of hourly forecasts while maintaining the dependency structure.

KW - Probabilistic electricity price forecasting

KW - scenario generation

KW - deep neural network

KW - non-parametric bayesian networks

KW - quantile regression

KW - probabilistic forecasting

KW - day ahead market

KW - demand response

UR - http://www.scopus.com/inward/record.url?scp=85142282431&partnerID=8YFLogxK

U2 - 10.1109/POWERCON53406.2022.9929940

DO - 10.1109/POWERCON53406.2022.9929940

M3 - Conference contribution

SN - 978-1-6654-1776-1

SP - 1

EP - 5

BT - Proceedings of the 2022 IEEE International Conference on Power Systems Technology (POWERCON)

PB - IEEE

Y2 - 12 September 2022 through 14 September 2022

ER -

van der Heijden T , Palensky P , van de Giesen N , Abraham E. Day Ahead Market price scenario generation using a Combined Quantile Regression Deep Neural Network and a Non-parametric Bayesian Network: A framework for risk-based Demand Response. In Proceedings of the 2022 IEEE International Conference on Power Systems Technology (POWERCON). IEEE. 2022. p. 1-5 doi: 10.1109/POWERCON53406.2022.9929940

Day Ahead Market price scenario generation using a Combined Quantile Regression Deep Neural Network and a Non-parametric Bayesian Network: A framework for risk-based Demand Response

Abstract

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Dataset underlying the PhD thesis 'Unlocking flexibility: risk-aware operational water and energy management'

Cite this

Day Ahead Market price scenario generation using a Combined Quantile Regression Deep Neural Network and a Non-parametric Bayesian Network: A framework for risk-based Demand Response

Abstract

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Datasets

Dataset underlying the PhD thesis 'Unlocking flexibility: risk-aware operational water and energy management'

Cite this