Adaptive transformer-based multi-task learning framework for synchronous prediction of substation flooding and outage risks

•A novel Transformer-based multi-task prediction model synchronously predicts flooding and outage risks within substations.•Adaptive time coding improves temporal dependency modeling for flooding and outage predictions.•Feature fusion strategy handles multivariate inputs and reduces redundant inform...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Electric power systems research Ročník 242; s. 111450
Hlavní autoři: Shi, Yu, Shi, Ying, Yao, Degui, Lu, Ming, Liang, Yun
Médium: Journal Article
Jazyk:angličtina
Vydáno: Elsevier B.V 01.05.2025
Témata:
ISSN:0378-7796
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Abstract •A novel Transformer-based multi-task prediction model synchronously predicts flooding and outage risks within substations.•Adaptive time coding improves temporal dependency modeling for flooding and outage predictions.•Feature fusion strategy handles multivariate inputs and reduces redundant information.•Cost-sensitive learning loss reduces the impact of imbalance on the data side, while joint-weighted loss reduces the impact of imbalance on the model training side.•A novel data-driven model offers reliable decision support for substation flooding prevention. Flooding disasters significantly threaten substation security, and forecasting risks of flooding and resulting outages within the substation is crucial for taking preventive measures and enhancing the substation's resilience. Existing models may suffer from low accuracy of risk prediction due to the difficulty of handling nonlinear multi-factors, dynamic temporal dependencies, and unbalanced data. Additionally, they rarely forecast flooding and outages simultaneously, leading to incomplete risk assessments. Therefore, a novel Transformer-based multi-task learning model (MTformer) is proposed to simultaneously predict flooding and outage risk within substations. MTformer is an attention-based shared encoder-decoder architecture that can achieve shared feature extraction and collaborative prediction. This model adopts three improved strategies: adaptive temporal encoding to enhance temporal dependency extraction, feature perception strategy to fuse heterogeneous data inputs, and training balancing strategy to balance multi-task training and reduce the impact of data imbalance. The experiment results show that the MTformer effectively predicts substation flooding and outage risks and outperforms the mainstream predictive model, with a decrease of 47.96 % in RMSE for flooding prediction and an increase of 39.82 % in F1 for outage prediction. Case studies demonstrate the potential of MTformer as a decision-making tool for proactive disaster mitigation and recovery planning.
AbstractList •A novel Transformer-based multi-task prediction model synchronously predicts flooding and outage risks within substations.•Adaptive time coding improves temporal dependency modeling for flooding and outage predictions.•Feature fusion strategy handles multivariate inputs and reduces redundant information.•Cost-sensitive learning loss reduces the impact of imbalance on the data side, while joint-weighted loss reduces the impact of imbalance on the model training side.•A novel data-driven model offers reliable decision support for substation flooding prevention. Flooding disasters significantly threaten substation security, and forecasting risks of flooding and resulting outages within the substation is crucial for taking preventive measures and enhancing the substation's resilience. Existing models may suffer from low accuracy of risk prediction due to the difficulty of handling nonlinear multi-factors, dynamic temporal dependencies, and unbalanced data. Additionally, they rarely forecast flooding and outages simultaneously, leading to incomplete risk assessments. Therefore, a novel Transformer-based multi-task learning model (MTformer) is proposed to simultaneously predict flooding and outage risk within substations. MTformer is an attention-based shared encoder-decoder architecture that can achieve shared feature extraction and collaborative prediction. This model adopts three improved strategies: adaptive temporal encoding to enhance temporal dependency extraction, feature perception strategy to fuse heterogeneous data inputs, and training balancing strategy to balance multi-task training and reduce the impact of data imbalance. The experiment results show that the MTformer effectively predicts substation flooding and outage risks and outperforms the mainstream predictive model, with a decrease of 47.96 % in RMSE for flooding prediction and an increase of 39.82 % in F1 for outage prediction. Case studies demonstrate the potential of MTformer as a decision-making tool for proactive disaster mitigation and recovery planning.
ArticleNumber 111450
Author Shi, Ying
Yao, Degui
Liang, Yun
Shi, Yu
Lu, Ming
Author_xml – sequence: 1
  givenname: Yu
  surname: Shi
  fullname: Shi, Yu
  organization: School of Automation, Wuhan University of Technology, Wuhan, Hubei 430070, China
– sequence: 2
  givenname: Ying
  orcidid: 0000-0003-3519-2308
  surname: Shi
  fullname: Shi, Ying
  email: a_laly@163.com
  organization: School of Automation, Wuhan University of Technology, Wuhan, Hubei 430070, China
– sequence: 3
  givenname: Degui
  surname: Yao
  fullname: Yao, Degui
  organization: State Grid Henan Electric Power Research Institute, Zhengzhou 450000, China
– sequence: 4
  givenname: Ming
  surname: Lu
  fullname: Lu, Ming
  organization: State Grid Henan Electric Power Research Institute, Zhengzhou 450000, China
– sequence: 5
  givenname: Yun
  surname: Liang
  fullname: Liang, Yun
  organization: State Grid Henan Electric Power Research Institute, Zhengzhou 450000, China
BookMark eNp9kMtOwzAQRb0oEm3hB1j5B1Js5y2xqSpeUiU2sLYce1zcJnZlO0UVP09CWLHoajTSnKu5Z4Fm1llA6I6SFSW0uN-v4Bj8ihGWryilWU5maE7SskrKsi6u0SKEPSGkqMt8jr7XShyjOQGOXtigne_AJ40IoHDXt9EkUYQDbkF4a-wOay86-HL-gIdTHM5WfnpnXR_w0YMyMhpnsdM49E2I4nfTrXNqZIVV2PVR7AB7Ew7hBl1p0Qa4_ZtL9PH0-L55SbZvz6-b9TaRKSExAdWUZZNRlkmWMkGkEHUNaZM1hJZasUqBSKlkmqWpKnOqi4LUVEnV1FUmqyZdomrKld6F4EFzaabfhs6m5ZTwURzf81EcH8XxSdyAsn_o0ZtO-PNl6GGCYCh1MuB5kAasHPx4kJErZy7hP8L1kB0
CitedBy_id crossref_primary_10_3390_ijgi14050189
crossref_primary_10_1109_JSTARS_2025_3596008
crossref_primary_10_1016_j_epsr_2025_112002
Cites_doi 10.1145/2347736.2347755
10.1109/TPAMI.2021.3054719
10.1613/jair.1.11192
10.1109/TMM.2019.2920613
10.1109/TSG.2022.3166600
10.3390/w10111536
10.1109/TPWRD.2016.2598572
10.1186/s40537-016-0043-6
10.1016/j.epsr.2024.110901
10.1109/TPWRS.2015.2429656
10.1016/j.ijforecast.2021.03.012
10.1016/j.epsr.2023.109473
10.1016/j.ress.2024.110169
10.1109/PESGM.2017.8273905
10.1038/nature14539
10.1016/j.epsr.2022.108098
10.1016/j.jhydrol.2020.124631
10.1038/s42256-022-00568-3
10.1016/j.comnet.2020.107744
10.1109/ACCESS.2024.3355484
10.9734/air/2023/v24i5961
10.1007/s00500-020-04954-0
10.1016/j.ijepes.2021.107545
10.1109/TIM.2023.3238059
10.2166/nh.2020.068
10.3390/su12041527
10.1109/TPWRS.2021.3086031
10.3390/en12020205
10.1016/j.ress.2022.108628
10.1109/TPWRS.2022.3146229
ContentType Journal Article
Copyright 2025 Elsevier B.V.
Copyright_xml – notice: 2025 Elsevier B.V.
DBID AAYXX
CITATION
DOI 10.1016/j.epsr.2025.111450
DatabaseName CrossRef
DatabaseTitle CrossRef
DatabaseTitleList
DeliveryMethod fulltext_linktorsrc
Discipline Engineering
ExternalDocumentID 10_1016_j_epsr_2025_111450
S0378779625000434
GroupedDBID --K
--M
-~X
.~1
0R~
1B1
1~.
1~5
29G
4.4
457
4G.
5GY
5VS
7-5
71M
8P~
9JN
AABNK
AAEDT
AAEDW
AAHCO
AAIKJ
AAKOC
AALRI
AAOAW
AAQFI
AAQXK
AARJD
AATTM
AAXKI
AAXUO
AAYWO
ABFNM
ABJNI
ABMAC
ABWVN
ABXDB
ACDAQ
ACGFS
ACIWK
ACNNM
ACRLP
ACRPL
ADBBV
ADEZE
ADHUB
ADMUD
ADNMO
ADTZH
AEBSH
AECPX
AEIPS
AEKER
AENEX
AFJKZ
AFTJW
AFXIZ
AGCQF
AGHFR
AGQPQ
AGRNS
AGUBO
AGYEJ
AHHHB
AHIDL
AHJVU
AI.
AIEXJ
AIIUN
AIKHN
AITUG
AKRWK
ALMA_UNASSIGNED_HOLDINGS
AMRAJ
ANKPU
APXCP
ARUGR
ASPBG
AVWKF
AXJTR
AZFZN
BELTK
BJAXD
BKOJK
BLXMC
BNPGV
CS3
DU5
E.L
EBS
EFJIC
EJD
EO8
EO9
EP2
EP3
FDB
FEDTE
FGOYB
FIRID
FNPLU
FYGXN
G-2
G-Q
GBLVA
HVGLF
HZ~
IHE
J1W
JARJE
JJJVA
K-O
KOM
LY6
LY7
M41
MO0
N9A
O-L
O9-
OAUVE
OZT
P-8
P-9
P2P
PC.
Q38
R2-
RIG
ROL
RPZ
SAC
SDF
SDG
SES
SET
SEW
SPC
SPCBC
SSH
SSR
SST
SSW
SSZ
T5K
VH1
WUQ
ZMT
~G-
9DU
AAYXX
ACLOT
ACVFH
ADCNI
AEUPX
AFPUW
AIGII
AKBMS
AKYEP
CITATION
EFKBS
EFLBG
~HD
ID FETCH-LOGICAL-c300t-edb77b4124c232a0caa99e3b4b017fd28dea31c2f233d751f66091dcdb984c8b3
ISICitedReferencesCount 3
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001421093100001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 0378-7796
IngestDate Tue Nov 18 22:23:08 EST 2025
Sat Nov 29 06:53:45 EST 2025
Sat May 24 17:07:14 EDT 2025
IsPeerReviewed true
IsScholarly true
Keywords Risk prediction and assessment
Transformer algorithm
Substation flooding
Multi-task learning
Substation outage
Language English
LinkModel OpenURL
MergedId FETCHMERGED-LOGICAL-c300t-edb77b4124c232a0caa99e3b4b017fd28dea31c2f233d751f66091dcdb984c8b3
ORCID 0000-0003-3519-2308
ParticipantIDs crossref_citationtrail_10_1016_j_epsr_2025_111450
crossref_primary_10_1016_j_epsr_2025_111450
elsevier_sciencedirect_doi_10_1016_j_epsr_2025_111450
PublicationCentury 2000
PublicationDate May 2025
2025-05-00
PublicationDateYYYYMMDD 2025-05-01
PublicationDate_xml – month: 05
  year: 2025
  text: May 2025
PublicationDecade 2020
PublicationTitle Electric power systems research
PublicationYear 2025
Publisher Elsevier B.V
Publisher_xml – name: Elsevier B.V
References Mosavi, Ozturk, Chau (bib0007) 2018; 10
Lim, Arık, Loeff, Pfister (bib0035) 2021; 37
Mohammadian, Aminifar, Amjady, Shahidehpour (bib0014) 2021; 36
Rizvi (bib0022) 2023; 24
Huang, Zhang, Chen, Feng, Li (bib0024) 2024; 12
Zhang, Shi, Yu, Yang, Lin (bib0030) 2023; 221
Fernandez, Garcia, Herrera, Chawla (bib0034) 2018; 61
Hou, Zhang, Wei, Huang, Liang, Li (bib0017) 2022; 210
Vaswani (bib0025) 2017
Xu, Zhao, Vineet, Lim, Torralba (bib0032) 2022
Duvnjak Žarković, Weiss, Hilber (bib0019) 2024; 236
Hou, Yu, Wang, Huang, Wu, Xu, Li, Geng (bib0018) 2019; 12
Zhou, Jiang, Wu, He, Weng, Peng (bib0031) 2019; 21
Hewage, Behera, Trovati, Pereira, Ghahremani, Palmieri, Liu (bib0040) 2020; 24
Huang, Jing, Li, Cai, Wu, Zhen (bib0013) 2017; 32
J.B. Thomas, S.G. Chaudhari, K.V. Shihabudheen, N.K. Verma, CNN-based transformer model for fault detection in power system networks, 72 (2023) 1–10.
Aerts, Botzen, Emanuel, Lin, de Moel, Michel-Kerjan (bib0010) 2014; 344
Movahednia, Kargarian, Ozdemir, Hagen (bib0003) 2021
Domingos (bib0020) 2012; 55
Wang, Chen, Wang, Baldick (bib0004) 2016; 31
Chen, Hui, Xie, Wan, Zhou, Pei (bib0006) 2021; 186
Ballard (bib0002) 2018
Miraee-Ashtiani, Vahedifard, Karimi-Ghartemani, Zhao, Mallakpour, AghaKouchak (bib0009) 2022; 37
Gal, Ghahramani (bib0036) 2016; 29
Zhou, Zhang, Peng, Zhang, Li, Xiong, Zhang (bib0041) 2021; 35
Souto, Yip, Wu, Austgen, Kutanoglu, Hasenbein, Yang, King, Santoso (bib0012) 2022; 135
Hughes, Watson, Cerrai, Zhang, Bagtzoglou, Zhang, Anagnostou (bib0016) 2024
He, Zhang, Ren, Sun (bib0037) 2016
Li, Gao, Zhou, Xu, Nils Roar Sælthun (bib0005) 2020; 51
Weiss, Khoshgoftaar, Wang (bib0039) 2016; 3
.
Boggess, Becker, Mitchell (bib0001) 2014
Khalil, Haq, ul Islam (bib0027) 2023; 228
Kao, Zhou, Chang, Chang (bib0023) 2020; 583
Richard Hauer Costa, G.R. McAllister, Substation flood program and flood hardening case study, (2017).
LeCun, Bengio, Hinton (bib0021) 2015; 521
Wang, Wang, Ding, Zheng, Hu, Zhang (bib0033) 2022; 13
van de Ven, Tuytelaars, Tolias (bib0038) 2022
Zhang, Yang (bib0028) 2017; 31
Sánchez-Muñoz, Domínguez-García, Martínez-Gomariz, Russo, Stevens, Pardo (bib0011) 2020; 12
Hughes, Zhang, Cerrai, Bagtzoglou, Wanik, Anagnostou (bib0015) 2022; 225
Vandenhende, Georgoulis, Van Gansbeke, Proesmans, Dai, Van Gool (bib0029) 2021
Movahednia (10.1016/j.epsr.2025.111450_bib0003) 2021
LeCun (10.1016/j.epsr.2025.111450_bib0021) 2015; 521
Rizvi (10.1016/j.epsr.2025.111450_bib0022) 2023; 24
Chen (10.1016/j.epsr.2025.111450_bib0006) 2021; 186
Zhang (10.1016/j.epsr.2025.111450_bib0030) 2023; 221
Souto (10.1016/j.epsr.2025.111450_bib0012) 2022; 135
Mohammadian (10.1016/j.epsr.2025.111450_bib0014) 2021; 36
Weiss (10.1016/j.epsr.2025.111450_bib0039) 2016; 3
Wang (10.1016/j.epsr.2025.111450_bib0004) 2016; 31
Hughes (10.1016/j.epsr.2025.111450_bib0015) 2022; 225
Hughes (10.1016/j.epsr.2025.111450_bib0016) 2024
Zhang (10.1016/j.epsr.2025.111450_bib0028) 2017; 31
van de Ven (10.1016/j.epsr.2025.111450_bib0038) 2022
Zhou (10.1016/j.epsr.2025.111450_bib0031) 2019; 21
10.1016/j.epsr.2025.111450_bib0026
Huang (10.1016/j.epsr.2025.111450_bib0024) 2024; 12
Hou (10.1016/j.epsr.2025.111450_bib0018) 2019; 12
He (10.1016/j.epsr.2025.111450_bib0037) 2016
Miraee-Ashtiani (10.1016/j.epsr.2025.111450_bib0009) 2022; 37
10.1016/j.epsr.2025.111450_bib0008
Aerts (10.1016/j.epsr.2025.111450_bib0010) 2014; 344
Xu (10.1016/j.epsr.2025.111450_bib0032) 2022
Kao (10.1016/j.epsr.2025.111450_bib0023) 2020; 583
Li (10.1016/j.epsr.2025.111450_bib0005) 2020; 51
Mosavi (10.1016/j.epsr.2025.111450_bib0007) 2018; 10
Domingos (10.1016/j.epsr.2025.111450_bib0020) 2012; 55
Khalil (10.1016/j.epsr.2025.111450_bib0027) 2023; 228
Fernandez (10.1016/j.epsr.2025.111450_bib0034) 2018; 61
Zhou (10.1016/j.epsr.2025.111450_bib0041) 2021; 35
Duvnjak Žarković (10.1016/j.epsr.2025.111450_bib0019) 2024; 236
Huang (10.1016/j.epsr.2025.111450_bib0013) 2017; 32
Sánchez-Muñoz (10.1016/j.epsr.2025.111450_bib0011) 2020; 12
Lim (10.1016/j.epsr.2025.111450_bib0035) 2021; 37
Ballard (10.1016/j.epsr.2025.111450_bib0002) 2018
Vaswani (10.1016/j.epsr.2025.111450_bib0025) 2017
Vandenhende (10.1016/j.epsr.2025.111450_bib0029) 2021
Wang (10.1016/j.epsr.2025.111450_bib0033) 2022; 13
Hewage (10.1016/j.epsr.2025.111450_bib0040) 2020; 24
Gal (10.1016/j.epsr.2025.111450_bib0036) 2016; 29
Boggess (10.1016/j.epsr.2025.111450_bib0001) 2014
Hou (10.1016/j.epsr.2025.111450_bib0017) 2022; 210
References_xml – volume: 12
  start-page: 15007
  year: 2024
  end-page: 15015
  ident: bib0024
  article-title: Prediction algorithm for power outage areas of affected customers based on CNNLSTM
  publication-title: IEEE Access.
– volume: 221
  year: 2023
  ident: bib0030
  article-title: Online measurement of capacitor voltage transformer metering errors based on GRU and MTL
  publication-title: Electr. Power Syst. Res.
– volume: 31
  start-page: 1604
  year: 2016
  end-page: 1613
  ident: bib0004
  article-title: Research on Resilience of power systems under natural disasters—a review
  publication-title: IEEE Trans. Power Syst.
– volume: 12
  start-page: 205
  year: 2019
  ident: bib0018
  article-title: Risk assessment and its visualization of power tower under typhoon disaster based on machine learning algorithms
  publication-title: Energies
– volume: 24
  start-page: 80
  year: 2023
  end-page: 88
  ident: bib0022
  article-title: Leveraging deep learning algorithms for predicting power outages and detecting faults: a review
  publication-title: Adv. Res.
– volume: 13
  start-page: 2703
  year: 2022
  end-page: 2714
  ident: bib0033
  article-title: A Transformer-based method of multienergy load forecasting in integrated energy system
  publication-title: IEEE Trans. Smart. Grid.
– start-page: 770
  year: 2016
  end-page: 778
  ident: bib0037
  article-title: Deep residual learning for image recognition
  publication-title: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
– volume: 225
  year: 2022
  ident: bib0015
  article-title: A hybrid physics-based and data-driven model for power distribution system infrastructure hardening and outage simulation
  publication-title: Reliab. Eng. Syst. Saf.
– start-page: 1
  year: 2021
  ident: bib0029
  article-title: Multi-task learning for dense prediction tasks: a survey
  publication-title: IEEE Trans. Pattern. Anal. Mach. Intell.
– volume: 21
  start-page: 3136
  year: 2019
  end-page: 3149
  ident: bib0031
  article-title: BranchGAN: unsupervised mutual image-to-image transfer with a single encoder and dual decoders
  publication-title: IEEE Trans. Multimedia
– volume: 583
  year: 2020
  ident: bib0023
  article-title: Exploring a long short-term memory based encoder-decoder framework for multi-step-ahead flood forecasting
  publication-title: J. Hydrol. (Amst)
– volume: 135
  year: 2022
  ident: bib0012
  article-title: Power system resilience to floods: Modeling, impact assessment, and mid-term mitigation strategies
  publication-title: Int. J. Electr. Power Energy Syst.
– volume: 228
  year: 2023
  ident: bib0027
  article-title: A deep learning-based transformer model for photovoltaic fault forecasting and classification
  publication-title: Electr. Power Syst. Res.
– year: 2017
  ident: bib0025
  article-title: Attention is all you need
  publication-title: Adv. Neural Inf. Process. Syst.
– volume: 12
  start-page: 1527
  year: 2020
  ident: bib0011
  article-title: Electrical grid risk assessment against flooding in Barcelona and Bristol cities
  publication-title: Sustainability
– year: 2022
  ident: bib0038
  article-title: Three types of incremental learning
  publication-title: Nat. Mach. Intell.
– volume: 344
  start-page: 473
  year: 2014
  end-page: 475
  ident: bib0010
  article-title: Evaluating flood resilience strategies for coastal megacities
  publication-title: Science (1979)
– start-page: 1
  year: 2014
  end-page: 5
  ident: bib0001
  article-title: Storm & flood hardening of electrical substations
  publication-title: IEEE Xplore
– volume: 36
  start-page: 4906
  year: 2021
  end-page: 4914
  ident: bib0014
  article-title: Data-driven classifier for extreme outage prediction based on bayes decision theory
  publication-title: IEEE Trans. Power Syst.
– volume: 51
  start-page: 366
  year: 2020
  end-page: 380
  ident: bib0005
  article-title: Usage of SIMWE model to model urban overland flood: a case study in Oslo
  publication-title: Hydrol. Res.
– volume: 10
  start-page: 1536
  year: 2018
  ident: bib0007
  article-title: Flood prediction using machine learning models: literature review
  publication-title: Water
– volume: 3
  year: 2016
  ident: bib0039
  article-title: A survey of transfer learning
  publication-title: J. Big. Data
– reference: Richard Hauer Costa, G.R. McAllister, Substation flood program and flood hardening case study, (2017).
– volume: 31
  year: 2017
  ident: bib0028
  article-title: Learning sparse task relations in multi-task learning
  publication-title: Proceedings of the AAAI Conference on Artificial Intelligence
– start-page: 304
  year: 2022
  end-page: 321
  ident: bib0032
  article-title: MTFormer: multi-task learning via transformer and cross-task reasoning
  publication-title: Lect. Notes Comput. Sci.
– volume: 210
  year: 2022
  ident: bib0017
  article-title: Review of failure risk and outage prediction in power system under wind hazards
  publication-title: Electr. Power Syst. Res.
– volume: 521
  start-page: 436
  year: 2015
  end-page: 444
  ident: bib0021
  article-title: Deep learning
  publication-title: Nature
– reference: J.B. Thomas, S.G. Chaudhari, K.V. Shihabudheen, N.K. Verma, CNN-based transformer model for fault detection in power system networks, 72 (2023) 1–10.
– volume: 236
  year: 2024
  ident: bib0019
  article-title: Addressing data deficiencies in outage reports: a qualitative and machine learning approach
  publication-title: Electr. Power Syst. Res.
– volume: 35
  start-page: 11106
  year: 2021
  end-page: 11115
  ident: bib0041
  article-title: Informer: beyond efficient transformer for long sequence time-series forecasting
  publication-title: Proceedings of the AAAI Conference on Artificial Intelligence
– year: 2018
  ident: bib0002
  article-title: Entergy seeks to charge many Louisiana customers to recoup nearly $6M lost amid 2016 flood
  publication-title: Advocate
– volume: 29
  year: 2016
  ident: bib0036
  article-title: A theoretically grounded application of dropout in recurrent neural networks
  publication-title: Neural Inf. Process. Syst.
– volume: 61
  start-page: 863
  year: 2018
  end-page: 905
  ident: bib0034
  article-title: SMOTE for learning from imbalanced data: progress and challenges, marking the 15-year anniversary
  publication-title: J. Artif. Intell. Res.
– reference: .
– volume: 186
  year: 2021
  ident: bib0006
  article-title: Convolutional neural networks for forecasting flood process in internet-of-things enabled smart city
  publication-title: Comput. Netw.
– volume: 37
  start-page: 4651
  year: 2022
  end-page: 4660
  ident: bib0009
  article-title: Performance degradation of levee-protected electric power network due to flooding in a changing climate
  publication-title: IEEE Trans. Power Syst.
– volume: 37
  start-page: 1748
  year: 2021
  end-page: 1764
  ident: bib0035
  article-title: Temporal Fusion Transformers for interpretable multi-horizon time series forecasting
  publication-title: Int. J. Forecast.
– volume: 24
  start-page: 16453
  year: 2020
  end-page: 16482
  ident: bib0040
  article-title: Temporal convolutional neural (TCN) network for an effective weather forecasting using time-series data from the local weather station
  publication-title: Soft. Comput.
– volume: 32
  start-page: 1098
  year: 2017
  end-page: 1105
  ident: bib0013
  article-title: Smart substation: state of the art and future development
  publication-title: IEEE Trans. Power Del.
– year: 2024
  ident: bib0016
  article-title: Assessing grid hardening strategies to improve power system performance during storms using a hybrid mechanistic-machine learning outage prediction model
  publication-title: Reliab. Eng. Syst. Saf.
– start-page: 1
  year: 2021
  ident: bib0003
  article-title: Power grid resilience enhancement via protecting electrical substations against flood hazards: a stochastic framework
  publication-title: IEEE Trans. Industr. Inform.
– volume: 55
  start-page: 78
  year: 2012
  end-page: 87
  ident: bib0020
  article-title: A few useful things to know about machine learning
  publication-title: Commun. ACM
– volume: 55
  start-page: 78
  year: 2012
  ident: 10.1016/j.epsr.2025.111450_bib0020
  article-title: A few useful things to know about machine learning
  publication-title: Commun. ACM
  doi: 10.1145/2347736.2347755
– year: 2017
  ident: 10.1016/j.epsr.2025.111450_bib0025
  article-title: Attention is all you need
  publication-title: Adv. Neural Inf. Process. Syst.
– start-page: 1
  year: 2021
  ident: 10.1016/j.epsr.2025.111450_bib0029
  article-title: Multi-task learning for dense prediction tasks: a survey
  publication-title: IEEE Trans. Pattern. Anal. Mach. Intell.
  doi: 10.1109/TPAMI.2021.3054719
– volume: 61
  start-page: 863
  year: 2018
  ident: 10.1016/j.epsr.2025.111450_bib0034
  article-title: SMOTE for learning from imbalanced data: progress and challenges, marking the 15-year anniversary
  publication-title: J. Artif. Intell. Res.
  doi: 10.1613/jair.1.11192
– volume: 344
  start-page: 473
  year: 2014
  ident: 10.1016/j.epsr.2025.111450_bib0010
  article-title: Evaluating flood resilience strategies for coastal megacities
  publication-title: Science (1979)
– volume: 21
  start-page: 3136
  year: 2019
  ident: 10.1016/j.epsr.2025.111450_bib0031
  article-title: BranchGAN: unsupervised mutual image-to-image transfer with a single encoder and dual decoders
  publication-title: IEEE Trans. Multimedia
  doi: 10.1109/TMM.2019.2920613
– volume: 13
  start-page: 2703
  year: 2022
  ident: 10.1016/j.epsr.2025.111450_bib0033
  article-title: A Transformer-based method of multienergy load forecasting in integrated energy system
  publication-title: IEEE Trans. Smart. Grid.
  doi: 10.1109/TSG.2022.3166600
– volume: 10
  start-page: 1536
  year: 2018
  ident: 10.1016/j.epsr.2025.111450_bib0007
  article-title: Flood prediction using machine learning models: literature review
  publication-title: Water
  doi: 10.3390/w10111536
– volume: 32
  start-page: 1098
  year: 2017
  ident: 10.1016/j.epsr.2025.111450_bib0013
  article-title: Smart substation: state of the art and future development
  publication-title: IEEE Trans. Power Del.
  doi: 10.1109/TPWRD.2016.2598572
– volume: 3
  year: 2016
  ident: 10.1016/j.epsr.2025.111450_bib0039
  article-title: A survey of transfer learning
  publication-title: J. Big. Data
  doi: 10.1186/s40537-016-0043-6
– volume: 236
  year: 2024
  ident: 10.1016/j.epsr.2025.111450_bib0019
  article-title: Addressing data deficiencies in outage reports: a qualitative and machine learning approach
  publication-title: Electr. Power Syst. Res.
  doi: 10.1016/j.epsr.2024.110901
– volume: 31
  start-page: 1604
  year: 2016
  ident: 10.1016/j.epsr.2025.111450_bib0004
  article-title: Research on Resilience of power systems under natural disasters—a review
  publication-title: IEEE Trans. Power Syst.
  doi: 10.1109/TPWRS.2015.2429656
– volume: 37
  start-page: 1748
  year: 2021
  ident: 10.1016/j.epsr.2025.111450_bib0035
  article-title: Temporal Fusion Transformers for interpretable multi-horizon time series forecasting
  publication-title: Int. J. Forecast.
  doi: 10.1016/j.ijforecast.2021.03.012
– volume: 221
  year: 2023
  ident: 10.1016/j.epsr.2025.111450_bib0030
  article-title: Online measurement of capacitor voltage transformer metering errors based on GRU and MTL
  publication-title: Electr. Power Syst. Res.
  doi: 10.1016/j.epsr.2023.109473
– year: 2024
  ident: 10.1016/j.epsr.2025.111450_bib0016
  article-title: Assessing grid hardening strategies to improve power system performance during storms using a hybrid mechanistic-machine learning outage prediction model
  publication-title: Reliab. Eng. Syst. Saf.
  doi: 10.1016/j.ress.2024.110169
– volume: 228
  year: 2023
  ident: 10.1016/j.epsr.2025.111450_bib0027
  article-title: A deep learning-based transformer model for photovoltaic fault forecasting and classification
  publication-title: Electr. Power Syst. Res.
– ident: 10.1016/j.epsr.2025.111450_bib0008
  doi: 10.1109/PESGM.2017.8273905
– volume: 521
  start-page: 436
  year: 2015
  ident: 10.1016/j.epsr.2025.111450_bib0021
  article-title: Deep learning
  publication-title: Nature
  doi: 10.1038/nature14539
– volume: 35
  start-page: 11106
  year: 2021
  ident: 10.1016/j.epsr.2025.111450_bib0041
  article-title: Informer: beyond efficient transformer for long sequence time-series forecasting
– year: 2018
  ident: 10.1016/j.epsr.2025.111450_bib0002
  article-title: Entergy seeks to charge many Louisiana customers to recoup nearly $6M lost amid 2016 flood
  publication-title: Advocate
– volume: 210
  year: 2022
  ident: 10.1016/j.epsr.2025.111450_bib0017
  article-title: Review of failure risk and outage prediction in power system under wind hazards
  publication-title: Electr. Power Syst. Res.
  doi: 10.1016/j.epsr.2022.108098
– volume: 583
  year: 2020
  ident: 10.1016/j.epsr.2025.111450_bib0023
  article-title: Exploring a long short-term memory based encoder-decoder framework for multi-step-ahead flood forecasting
  publication-title: J. Hydrol. (Amst)
  doi: 10.1016/j.jhydrol.2020.124631
– start-page: 304
  year: 2022
  ident: 10.1016/j.epsr.2025.111450_bib0032
  article-title: MTFormer: multi-task learning via transformer and cross-task reasoning
  publication-title: Lect. Notes Comput. Sci.
– year: 2022
  ident: 10.1016/j.epsr.2025.111450_bib0038
  article-title: Three types of incremental learning
  publication-title: Nat. Mach. Intell.
  doi: 10.1038/s42256-022-00568-3
– volume: 186
  year: 2021
  ident: 10.1016/j.epsr.2025.111450_bib0006
  article-title: Convolutional neural networks for forecasting flood process in internet-of-things enabled smart city
  publication-title: Comput. Netw.
  doi: 10.1016/j.comnet.2020.107744
– volume: 12
  start-page: 15007
  year: 2024
  ident: 10.1016/j.epsr.2025.111450_bib0024
  article-title: Prediction algorithm for power outage areas of affected customers based on CNNLSTM
  publication-title: IEEE Access.
  doi: 10.1109/ACCESS.2024.3355484
– volume: 24
  start-page: 80
  year: 2023
  ident: 10.1016/j.epsr.2025.111450_bib0022
  article-title: Leveraging deep learning algorithms for predicting power outages and detecting faults: a review
  publication-title: Adv. Res.
  doi: 10.9734/air/2023/v24i5961
– volume: 24
  start-page: 16453
  year: 2020
  ident: 10.1016/j.epsr.2025.111450_bib0040
  article-title: Temporal convolutional neural (TCN) network for an effective weather forecasting using time-series data from the local weather station
  publication-title: Soft. Comput.
  doi: 10.1007/s00500-020-04954-0
– volume: 135
  year: 2022
  ident: 10.1016/j.epsr.2025.111450_bib0012
  article-title: Power system resilience to floods: Modeling, impact assessment, and mid-term mitigation strategies
  publication-title: Int. J. Electr. Power Energy Syst.
  doi: 10.1016/j.ijepes.2021.107545
– ident: 10.1016/j.epsr.2025.111450_bib0026
  doi: 10.1109/TIM.2023.3238059
– volume: 51
  start-page: 366
  year: 2020
  ident: 10.1016/j.epsr.2025.111450_bib0005
  article-title: Usage of SIMWE model to model urban overland flood: a case study in Oslo
  publication-title: Hydrol. Res.
  doi: 10.2166/nh.2020.068
– start-page: 1
  year: 2014
  ident: 10.1016/j.epsr.2025.111450_bib0001
  article-title: Storm & flood hardening of electrical substations
  publication-title: IEEE Xplore
– start-page: 1
  year: 2021
  ident: 10.1016/j.epsr.2025.111450_bib0003
  article-title: Power grid resilience enhancement via protecting electrical substations against flood hazards: a stochastic framework
  publication-title: IEEE Trans. Industr. Inform.
– volume: 12
  start-page: 1527
  year: 2020
  ident: 10.1016/j.epsr.2025.111450_bib0011
  article-title: Electrical grid risk assessment against flooding in Barcelona and Bristol cities
  publication-title: Sustainability
  doi: 10.3390/su12041527
– volume: 36
  start-page: 4906
  year: 2021
  ident: 10.1016/j.epsr.2025.111450_bib0014
  article-title: Data-driven classifier for extreme outage prediction based on bayes decision theory
  publication-title: IEEE Trans. Power Syst.
  doi: 10.1109/TPWRS.2021.3086031
– volume: 31
  year: 2017
  ident: 10.1016/j.epsr.2025.111450_bib0028
  article-title: Learning sparse task relations in multi-task learning
– volume: 12
  start-page: 205
  year: 2019
  ident: 10.1016/j.epsr.2025.111450_bib0018
  article-title: Risk assessment and its visualization of power tower under typhoon disaster based on machine learning algorithms
  publication-title: Energies
  doi: 10.3390/en12020205
– volume: 225
  year: 2022
  ident: 10.1016/j.epsr.2025.111450_bib0015
  article-title: A hybrid physics-based and data-driven model for power distribution system infrastructure hardening and outage simulation
  publication-title: Reliab. Eng. Syst. Saf.
  doi: 10.1016/j.ress.2022.108628
– volume: 29
  year: 2016
  ident: 10.1016/j.epsr.2025.111450_bib0036
  article-title: A theoretically grounded application of dropout in recurrent neural networks
  publication-title: Neural Inf. Process. Syst.
– volume: 37
  start-page: 4651
  year: 2022
  ident: 10.1016/j.epsr.2025.111450_bib0009
  article-title: Performance degradation of levee-protected electric power network due to flooding in a changing climate
  publication-title: IEEE Trans. Power Syst.
  doi: 10.1109/TPWRS.2022.3146229
– start-page: 770
  year: 2016
  ident: 10.1016/j.epsr.2025.111450_bib0037
  article-title: Deep residual learning for image recognition
SSID ssj0006975
Score 2.454177
Snippet •A novel Transformer-based multi-task prediction model synchronously predicts flooding and outage risks within substations.•Adaptive time coding improves...
SourceID crossref
elsevier
SourceType Enrichment Source
Index Database
Publisher
StartPage 111450
SubjectTerms Multi-task learning
Risk prediction and assessment
Substation flooding
Substation outage
Transformer algorithm
Title Adaptive transformer-based multi-task learning framework for synchronous prediction of substation flooding and outage risks
URI https://dx.doi.org/10.1016/j.epsr.2025.111450
Volume 242
WOSCitedRecordID wos001421093100001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVESC
  databaseName: Elsevier SD Freedom Collection Journals 2021
  issn: 0378-7796
  databaseCode: AIEXJ
  dateStart: 19950101
  customDbUrl:
  isFulltext: true
  dateEnd: 99991231
  titleUrlDefault: https://www.sciencedirect.com
  omitProxy: false
  ssIdentifier: ssj0006975
  providerName: Elsevier
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1LT9wwELa2Sw_toepTUGjlA7dVqo3zcHxcISpaIdQDVNtT5FfQUpSNNgmi6l_oj-44tpNlaREcuESR5Xijnc8zY-fzNwjtp0yGRIawOolhNsVTLgIeyySAYKWooGmRdeXbvh_Tk5NsPmffRqM__izM1SUty-z6mlWPampoA2Obo7MPMHc_KDTAPRgdrmB2uN7L8DPFq44P1PicVK8CE6yUZQ8GDa9_-moR55PCs7M6wmH9q5RGLtcQY6uV-YjjM8oaPIxnJhqyuz_cuGwbQ_sxFPX6xjZ_V19nISeVqcPmFKPNN4q17bNOGrKjE_xoNxt8SDUeiS-tazxvF77tuLWkf9fL7VuQZGAJ-vNasH6l1Naz9b6YWKkt503BD8dWlvaWo7d7DhcQs2uj6kqST0Pnm6raG9Gu5yB6ettFbsbIzRi5HeMJ2iI0YdkYbc2-HM6_9pE9ZZ1wc__m7hCW5Qtuvsm_E5215OX0JXrhVh14ZtHyCo10-Ro9X9OifIN-e9zgW7jBA26wxw3ucYOhK17DDR5wg5cFHnCDPW4w4AZb3OAON2_R2efD04OjwFXmCGQ0nTaBVoJSYQqXS8jI-VRyzpiORCzAwReKZErzKJSkIBFM-SQs0hTyUiWVYFksMxG9Q-NyWepthEVasEJCFk-1iomAgWK4hVWBFBxyd72DQv835tLJ1pvqKZf5_w24gyb9M5UVbbmzd-Ktk7u006aTOYDtjufeP-hXdtGzYRbsoXGzavUH9FReNYt69dEh7S9idKpk
linkProvider Elsevier
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Adaptive+transformer-based+multi-task+learning+framework+for+synchronous+prediction+of+substation+flooding+and+outage+risks&rft.jtitle=Electric+power+systems+research&rft.au=Shi%2C+Yu&rft.au=Shi%2C+Ying&rft.au=Yao%2C+Degui&rft.au=Lu%2C+Ming&rft.date=2025-05-01&rft.issn=0378-7796&rft.volume=242&rft.spage=111450&rft_id=info:doi/10.1016%2Fj.epsr.2025.111450&rft.externalDBID=n%2Fa&rft.externalDocID=10_1016_j_epsr_2025_111450
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0378-7796&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0378-7796&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0378-7796&client=summon