基于Transformer的局部时间步长二维水动力方法研究

TV122%TV131; 提出了一种基于Transformer的局部时间步长二维水动力模拟方法,旨在优化洪涝模拟的速度与精度,满足迫切的防洪减灾需求.该方法通过引入Transformer模型的自注意力机制与序列预测功能,结合局部时间步长技术动态调整水动力系统中局部区域的时间步长.以蒙洼蓄洪区的实际洪水过程为实例,模拟结果表明:在与遥感淹没影像及传统数值方法模拟结果的对比中,该方法在淹没范围与淹没深度的预测方面展现出较高的一致性和可靠性,相对误差保持在4％之内;与传统的全局时间步长方法相比,该方法能在保证模拟稳定性和精度的前提下,显著减少计算时间,在150 m网格分辨率下计算效率提升70.27％...

Full description

Saved in:

Bibliographic Details
Published in	中国农村水利水电 no. 12; pp. 74 - 89
Main Authors	滕述臣, 向小华, 曹红伟, 林显, 吴晓玲
Format	Journal Article
Language	Chinese
Published	河海大学水文水资源学院,江苏南京 210098%中国长江电力股份有限公司,湖北宜昌 443002 15.12.2024
Subjects	Transformer hydrodynamic simulation 局部时间步长水动力模拟 time series encoding 时间序列编码 local time stepping
Online Access	Get full text
ISSN	1007-2284
DOI	10.12396/znsd.240621

Cover

Abstract	TV122%TV131; 提出了一种基于Transformer的局部时间步长二维水动力模拟方法,旨在优化洪涝模拟的速度与精度,满足迫切的防洪减灾需求.该方法通过引入Transformer模型的自注意力机制与序列预测功能,结合局部时间步长技术动态调整水动力系统中局部区域的时间步长.以蒙洼蓄洪区的实际洪水过程为实例,模拟结果表明:在与遥感淹没影像及传统数值方法模拟结果的对比中,该方法在淹没范围与淹没深度的预测方面展现出较高的一致性和可靠性,相对误差保持在4％之内;与传统的全局时间步长方法相比,该方法能在保证模拟稳定性和精度的前提下,显著减少计算时间,在150 m网格分辨率下计算效率提升70.27％.
AbstractList	TV122%TV131; 提出了一种基于Transformer的局部时间步长二维水动力模拟方法,旨在优化洪涝模拟的速度与精度,满足迫切的防洪减灾需求.该方法通过引入Transformer模型的自注意力机制与序列预测功能,结合局部时间步长技术动态调整水动力系统中局部区域的时间步长.以蒙洼蓄洪区的实际洪水过程为实例,模拟结果表明:在与遥感淹没影像及传统数值方法模拟结果的对比中,该方法在淹没范围与淹没深度的预测方面展现出较高的一致性和可靠性,相对误差保持在4％之内;与传统的全局时间步长方法相比,该方法能在保证模拟稳定性和精度的前提下,显著减少计算时间,在150 m网格分辨率下计算效率提升70.27％.
Abstract_FL	This paper proposes a Transformer-based local timestep two-dimensional hydrodynamic simulation method aimed at optimizing the speed and accuracy of flood simulations to meet the urgent needs of flood prevention and mitigation. By incorporating the self-attention mechanism and sequence prediction capabilities of the Transformer model,combined with local timestep techniques,this method dynamical-ly adjusts the local timestep in the hydrodynamic system. Using the real flood process in the Mengwa flood storage area as an example,the simulation results demonstrate that,compared to remote sensing inundation imagery and traditional numerical methods,this approach shows higher consistency and reliability in the precise prediction of inundation extent and depth,maintaining a relative error within 4％. Compared to traditional global timestep methods,this method significantly reduces computational time while ensuring simulation stability and accura-cy,achieving a 70.27％ improvement in computational efficiency at a 150 m grid resolution.
Author	滕述臣林显吴晓玲曹红伟向小华
AuthorAffiliation	河海大学水文水资源学院,江苏南京 210098%中国长江电力股份有限公司,湖北宜昌 443002
AuthorAffiliation_xml	– name: 河海大学水文水资源学院,江苏南京 210098%中国长江电力股份有限公司,湖北宜昌 443002
Author_FL	XIANG Xiao-hua TENG Shu-chen WU Xiao-ling CAO Hong-wei LIN Xian
Author_FL_xml	– sequence: 1 fullname: TENG Shu-chen – sequence: 2 fullname: XIANG Xiao-hua – sequence: 3 fullname: CAO Hong-wei – sequence: 4 fullname: LIN Xian – sequence: 5 fullname: WU Xiao-ling
Author_xml	– sequence: 1 fullname: 滕述臣 – sequence: 2 fullname: 向小华 – sequence: 3 fullname: 曹红伟 – sequence: 4 fullname: 林显 – sequence: 5 fullname: 吴晓玲
BookMark	eNotjT9Lw0AYh2-oYK3d_BIOqfe-l8tdx1L8BwWXOpfkkitKvUAOEToJOhVxshZ00EWog6BYCsbBL9Mk5FsY0OkHDw_Pb4PUTGwiQraAtgBZ29sZGxu20KUeQo3UgVLhIEp3nTStPQkohQq1BdRJJ3tKV-ltP_GN1XFyFiXFw3X2cVlezfPZspwt8reXcvqzSm-K70X-vsgm82zymN9_5Z_T4vmueF1ukjXtj2zU_N8GOd7b7XcPnN7R_mG303MsUI86QqBkAfqMA9NcMSWk4opHbohcCNAaJRdhpQaBVFQy5OCBL0BIJpkKPNYg23_dC99o3wwHp_F5YqrHwXholB3ZECm6gBQo-wWZ-2CZ
ClassificationCodes	TV122%TV131
ContentType	Journal Article
Copyright	Copyright © Wanfang Data Co. Ltd. All Rights Reserved.
Copyright_xml	– notice: Copyright © Wanfang Data Co. Ltd. All Rights Reserved.
DBID	2B. 4A8 92I 93N PSX TCJ
DOI	10.12396/znsd.240621
DatabaseName	Wanfang Data Journals - Hong Kong WANFANG Data Centre Wanfang Data Journals 万方数据期刊 - 香港版 China Online Journals (COJ) China Online Journals (COJ)
DatabaseTitleList
DeliveryMethod	fulltext_linktorsrc
Discipline	Engineering
DocumentTitle_FL	Research on the Local Time-Step Two-Dimensional Hydrodynamic Method Based on Transformer
EndPage	89
ExternalDocumentID	zgncslsd202412010
GrantInformation_xml	– fundername: (国家重点研发计划); (中国长江电力股份公司技术服务项目); (中国长江电力股份公司技术服务项目) funderid: (国家重点研发计划); (中国长江电力股份公司技术服务项目); (中国长江电力股份公司技术服务项目)
GroupedDBID	-04 2B. 4A8 5XA 5XD 92H 92I 93N ABJNI ACGFS ALMA_UNASSIGNED_HOLDINGS CCEZO CHDYS CW9 GROUPED_DOAJ PSX TCJ TGT U1G U5M
ID	FETCH-LOGICAL-s1060-77283b2a3513f5c3c78c5c5e4d25771ff2857d106bb8c08325161a7178383cb63
ISSN	1007-2284
IngestDate	Thu May 29 04:09:01 EDT 2025
IsPeerReviewed	false
IsScholarly	true
Issue	12
Keywords	Transformer hydrodynamic simulation 局部时间步长水动力模拟 time series encoding 时间序列编码 local time stepping
Language	Chinese
LinkModel	OpenURL
MergedId	FETCHMERGED-LOGICAL-s1060-77283b2a3513f5c3c78c5c5e4d25771ff2857d106bb8c08325161a7178383cb63
PageCount	16
ParticipantIDs	wanfang_journals_zgncslsd202412010
PublicationCentury	2000
PublicationDate	2024-12-15
PublicationDateYYYYMMDD	2024-12-15
PublicationDate_xml	– month: 12 year: 2024 text: 2024-12-15 day: 15
PublicationDecade	2020
PublicationTitle	中国农村水利水电
PublicationTitle_FL	China Rural Water and Hydropower
PublicationYear	2024
Publisher	河海大学水文水资源学院,江苏南京 210098%中国长江电力股份有限公司,湖北宜昌 443002
Publisher_xml	– name: 河海大学水文水资源学院,江苏南京 210098%中国长江电力股份有限公司,湖北宜昌 443002
SSID	ssib001100971 ssj0037555 ssib051368504 ssib046786273
Score	2.3753357
Snippet	TV122%TV131;...
SourceID	wanfang
SourceType	Aggregation Database
StartPage	74
Title	基于Transformer的局部时间步长二维水动力方法研究
URI	https://d.wanfangdata.com.cn/periodical/zgncslsd202412010
hasFullText	1
inHoldings	1
isFullTextHit
isPrint
link	http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwnV07b9RAELai0ECBeIq3ImCryMH78u6WtuNTREGVSOmis30XCnRIXNJchQRVhKgIkaCABikUSCCiSISCP5O7U_4FM2P7zoEUgcaaG89-8_Jjds-763n3ddDGP4uUn6vc-LjkuZ8V3PpBB-pfzXN4HtJqn4_CpRX1cFWvzszea3y1tLmRLeSDE-eV_E9WgQd5xVmy_5DZCSgwgIb8whEyDMdT5ZilmrkWiyOWKjzadLmuQyETqWEOeAqlYo6fNKSOWckiy9KQOcPiEDlIKOREiyzSxAH5Vg2ZIE4cVzJxQIRmNiIcIlxMgHDWkYxEBGgVBcwpIhzoapbBBG5RI3oA4ERYAEkIapE5flydRZApxyByrOvLhc7EpNai7SAFhAXNcioCmgLC1Xjetgh3EYLWREFjHLkMDgqyM4EYHxNJMWZIgK60OWwiaHHGcuIoXehkl8CIIAF6DSqNFItKAiIeNryiIFrT4FhsZcvQR2j_tBXkyaGJgkIGCYYrAZ2Op66hmZTFKJmHTncABgt9UuwnOZ8AVQFuZBigBYs4NS9bUbqsq0wpjbNgX0JEC7VUxjl0DJkJ2QQyaZ1qgE3mlZJB0BiAxo9qfCHKbf3qZ4NovAHLPZfqWsqd-JYW0uGI0aDXLxawoiznyP-x7vlgvZf3n_QLTB4XNI3yjDCm_BajGjehmp_TOmj1b3jxQyd9WpNrjlss4JhDWZ5Jo2kr5Ikj1WwYtOlBwyKaxNfrtnvrjXpz-YJ3vuoozkXlXX_Rmxk8vuSdaywfetmLhh8ODg9eN-758buXw2_Pj17sjnb2j3b2Rl8-HW3_Ojx4Nf65N_q6N9zaHW69H739Mfq-Pf74Zvx5_4q30kqXkyW_2hLF7_MgDLAvbGUm2hLc6upc5sbmOtcdVcCr1_BuV1htChDNMptD7wp6LyFvG26stDLPQnnVm-097XWueXOuyI3MOly5rlSQ5jaOfHQFdIicKqAfcd27WwVgrXrk9df-ysmN0wjd9M5O78Bb3uzGs83ObSjlN7I7lMrfQ6GtuQ
linkProvider	Directory of Open Access Journals
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=%E5%9F%BA%E4%BA%8ETransformer%E7%9A%84%E5%B1%80%E9%83%A8%E6%97%B6%E9%97%B4%E6%AD%A5%E9%95%BF%E4%BA%8C%E7%BB%B4%E6%B0%B4%E5%8A%A8%E5%8A%9B%E6%96%B9%E6%B3%95%E7%A0%94%E7%A9%B6&rft.jtitle=%E4%B8%AD%E5%9B%BD%E5%86%9C%E6%9D%91%E6%B0%B4%E5%88%A9%E6%B0%B4%E7%94%B5&rft.au=%E6%BB%95%E8%BF%B0%E8%87%A3&rft.au=%E5%90%91%E5%B0%8F%E5%8D%8E&rft.au=%E6%9B%B9%E7%BA%A2%E4%BC%9F&rft.au=%E6%9E%97%E6%98%BE&rft.date=2024-12-15&rft.pub=%E6%B2%B3%E6%B5%B7%E5%A4%A7%E5%AD%A6%E6%B0%B4%E6%96%87%E6%B0%B4%E8%B5%84%E6%BA%90%E5%AD%A6%E9%99%A2%2C%E6%B1%9F%E8%8B%8F%E5%8D%97%E4%BA%AC+210098%25%E4%B8%AD%E5%9B%BD%E9%95%BF%E6%B1%9F%E7%94%B5%E5%8A%9B%E8%82%A1%E4%BB%BD%E6%9C%89%E9%99%90%E5%85%AC%E5%8F%B8%2C%E6%B9%96%E5%8C%97%E5%AE%9C%E6%98%8C+443002&rft.issn=1007-2284&rft.issue=12&rft.spage=74&rft.epage=89&rft_id=info:doi/10.12396%2Fznsd.240621&rft.externalDocID=zgncslsd202412010
thumbnail_s	http://utb.summon.serialssolutions.com/2.0.0/image/custom?url=http%3A%2F%2Fwww.wanfangdata.com.cn%2Fimages%2FPeriodicalImages%2Fzgncslsd%2Fzgncslsd.jpg