An advanced generative deep learning framework for probabilistic spatio-temporal wind power forecasting

dc.contributor.authorJalali, S. M.
dc.contributor.authorKhodayar, M.
dc.contributor.authorKhosravi, A.
dc.contributor.authorNahavandi, S.
dc.contributor.authorCatalão, João P. S.
dc.contributor.authorOsório, Gerardo J.
dc.date.accessioned2022-02-01T12:41:29Z
dc.date.available2022-02-01T12:41:29Z
dc.date.issued2021-09
dc.description.abstractThis paper presents a deep generative model for capturing the conditional probability distribution of future wind power given its history by modeling and pattern recognition in a dynamic graph. The dynamic nodes show the wind sites while the dynamic edges reflect the correlation between the nodes. We propose a scalable optimization model, which is theoretically proved to catch distributions at nodes of the graph, contrary to all learning formulations in the sector of discriminatory pattern recognition. The density of probabilities for each node can be used as samples in our framework. This probabilistic deep convolutional Auto-encoder (PDCA), is based on the deep learning of localized first-order approximation of spectral graph convolutions, a novel evolutionary algorithm, and the Bayesian variational inference concepts. The presented generative model is used for the spatio-temporal probabilistic wind power problem in a wide 25 wind sites located in California, the USA for up to 24h ahead prediction. The experimental findings reveal that our proposed model outperforms other competitive temporal and spatio-temporal algorithms in terms of reliability, sharpness, and continuously ranked probability score.pt_PT
dc.identifier.citationJalali, S. M., Khodayar, M, Khosravi, A., Osório, G. J., Nahavandi, S., & Catalão, J. P. S. (2021). An advanced generative deep learning framework for probabilistic spatio-temporal wind power forecasting. In Proceedings of the 21th IEEE International Conference on Environment and Electrical Engineering and 5th IEEE Industrial and Commercial Power Systems Europe (EEEIC 2021 / I&CPS Europe 2021), Bari, Italy, 7-10 September 2021 (pp. 1-6). doi: 10.1109/EEEIC/ICPSEurope51590.2021.9584664. Disponível no Repositório UPT, http://hdl.handle.net/11328/3920pt_PT
dc.identifier.doi10.1109/EEEIC/ICPSEurope51590.2021.9584664pt_PT
dc.identifier.isbn978-1-6654-3613-7
dc.identifier.urihttp://hdl.handle.net/11328/3920
dc.language.isoengpt_PT
dc.peerreviewedyespt_PT
dc.publisherIEEEpt_PT
dc.relation.publisherversionhttps://doi.org/10.1109/EEEIC/ICPSEurope51590.2021.9584664pt_PT
dc.rightsrestricted accesspt_PT
dc.rights.urihttp://creativecommons.org/licenses/by/4.0/pt_PT
dc.subjectDeep learningpt_PT
dc.subjectProbabilistic forecastingpt_PT
dc.subjectVariational bayesian inferencept_PT
dc.subjectSpectral graph convulutionspt_PT
dc.subjectEvolutionary algorithmpt_PT
dc.titleAn advanced generative deep learning framework for probabilistic spatio-temporal wind power forecastingpt_PT
dc.typeconferenceObjectpt_PT
degois.publication.firstPage1pt_PT
degois.publication.lastPage6pt_PT
degois.publication.locationBari, Italy, 7-10 September, 2021pt_PT
degois.publication.titleProceedings of the 21th IEEE International Conference on Environment and Electrical Engineering and 5th IEEE Industrial and Commercial Power Systems Europe (EEEIC 2021 / I&CPS Europe 2021)pt_PT
dspace.entity.typePublicationen
person.affiliation.nameREMIT – Research on Economics, Management and Information Technologies
person.familyNameOsório
person.givenNameGerardo J.
person.identifier.ciencia-idBD19-D0AD-65DB
person.identifier.gsidt13DoaMAAAAJ
person.identifier.orcid0000-0001-8328-9708
person.identifier.ridC-3616-2014
person.identifier.scopus-author-id54783251300
relation.isAuthorOfPublication7ce5da40-610d-4361-a87f-5cbdfe392256
relation.isAuthorOfPublication.latestForDiscovery7ce5da40-610d-4361-a87f-5cbdfe392256

Files

Original bundle

Now showing 1 - 1 of 1
Name:
9584664.pdf
Size:
1.72 MB
Format:
Adobe Portable Document Format