File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Article: Probabilistic individual load forecasting using pinball loss guided LSTM

TitleProbabilistic individual load forecasting using pinball loss guided LSTM
Authors
KeywordsDemand response
Individual consumer
Long short-term memory (LSTM)
Pinball loss
Probabilistic load forecasting
Quantile regression
Smart meter
Issue Date2019
Citation
Applied Energy, 2019, v. 235, p. 10-20 How to Cite?
AbstractThe installation of smart meters enables the collection of massive fine-grained electricity consumption data and makes individual consumer level load forecasting possible. Compared to aggregated loads, load forecasting for individual consumers is prone to non-stationary and stochastic features. In this paper, a probabilistic load forecasting method for individual consumers is proposed to handle the variability and uncertainty of future load profiles. Specifically, a deep neural network, long short-term memory (LSTM), is used to model both the long-term and short-term dependencies within the load profiles. Pinball loss, instead of the mean square error (MSE), is used to guide the training of the parameters. In this way, traditional LSTM-based point forecasting is extended to probabilistic forecasting in the form of quantiles. Numerical experiments are conducted on an open dataset from Ireland. Forecasting for both residential and commercial consumers is tested. Results show that the proposed method has superior performance over traditional methods.
Persistent Identifierhttp://hdl.handle.net/10722/308770
ISSN
2023 Impact Factor: 10.1
2023 SCImago Journal Rankings: 2.820
ISI Accession Number ID

 

DC FieldValueLanguage
dc.contributor.authorWang, Yi-
dc.contributor.authorGan, Dahua-
dc.contributor.authorSun, Mingyang-
dc.contributor.authorZhang, Ning-
dc.contributor.authorLu, Zongxiang-
dc.contributor.authorKang, Chongqing-
dc.date.accessioned2021-12-08T07:50:05Z-
dc.date.available2021-12-08T07:50:05Z-
dc.date.issued2019-
dc.identifier.citationApplied Energy, 2019, v. 235, p. 10-20-
dc.identifier.issn0306-2619-
dc.identifier.urihttp://hdl.handle.net/10722/308770-
dc.description.abstractThe installation of smart meters enables the collection of massive fine-grained electricity consumption data and makes individual consumer level load forecasting possible. Compared to aggregated loads, load forecasting for individual consumers is prone to non-stationary and stochastic features. In this paper, a probabilistic load forecasting method for individual consumers is proposed to handle the variability and uncertainty of future load profiles. Specifically, a deep neural network, long short-term memory (LSTM), is used to model both the long-term and short-term dependencies within the load profiles. Pinball loss, instead of the mean square error (MSE), is used to guide the training of the parameters. In this way, traditional LSTM-based point forecasting is extended to probabilistic forecasting in the form of quantiles. Numerical experiments are conducted on an open dataset from Ireland. Forecasting for both residential and commercial consumers is tested. Results show that the proposed method has superior performance over traditional methods.-
dc.languageeng-
dc.relation.ispartofApplied Energy-
dc.subjectDemand response-
dc.subjectIndividual consumer-
dc.subjectLong short-term memory (LSTM)-
dc.subjectPinball loss-
dc.subjectProbabilistic load forecasting-
dc.subjectQuantile regression-
dc.subjectSmart meter-
dc.titleProbabilistic individual load forecasting using pinball loss guided LSTM-
dc.typeArticle-
dc.description.naturelink_to_subscribed_fulltext-
dc.identifier.doi10.1016/j.apenergy.2018.10.078-
dc.identifier.scopuseid_2-s2.0-85055915128-
dc.identifier.volume235-
dc.identifier.spage10-
dc.identifier.epage20-
dc.identifier.isiWOS:000458942800002-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats