基于深度學(xué)習(xí)模型的混凝土壩變形預(yù)測研究

打開文本圖片集

關(guān)鍵詞：混凝土壩；變形監(jiān)測；變形預(yù)測；LSTM；TCN中圖分類號(hào)：TV642 文獻(xiàn)標(biāo)志碼：A doi：10.3969/j.issn.1000-1379.2025.07.023引用格式：，，，等.基于深度學(xué)習(xí)模型的混凝土壩變形預(yù)測研究[J].人民黃河，2025，47（7）：144-149，155.

Research on Deformation Prediction of Concrete Dams Based on Deep Learning Models

SU Xiaojun1，XU Zengguang1， ZHANG Ye1，KANG Xinyu'， ZHOU Tao2，YANG Tao3，LI Kangping4 （1.State KeyLaboratoryof WaterEngineering EcologyandEnvironment inAridArea，Xi’an Universityof Technology，Xi'an 710048，China；2.Huanghe Hydropower Development Co.，Ltd.， Xining 81000，China； 3.China Yangtze Power Co.，Ltd.，Yibin 644612，China；4.Power China Northwest Engineering Co.，Ltd.，Xi'an 710065，China） Abstract： Inordertoimprovetheacuracyofpredictingconcretedamdeformationsandensuredamstructuralsafety，thispaperproposeda predictionmodelcombiningTemporalConvolutionalNetworks（TCN）andLongShort-TermMemory（LSTM）networks，imingthelimitationsofexistingpredictionmodelsincaptuingcomplexdeformationfeaturesandlong-tedependencies.TCNwasusedtoextractteporal featuresfromdeformationdta，hileLSTMcapturedlongtedependenciesiscombinationsignificantlyehancedteodel’sabilityo predictomplexdeformationpatesExperimentalesultsemonstratetat，compaedwithvariousinteligentpedictionodels，eo posedmodelaccuratelysiulatestdadfaprocsndcheshigerpredictiouacyFurteore，eodelosates asignificantadvantageincapturigomplexdefoationcharacteisticsofconcretedams，providingaovelteciquefordamsafetoi ring.

Key words：concrete dam；deformation monitoring；deformation prediction；LSTM；TCN

0 引言

水庫大壩作為重要的基礎(chǔ)設(shè)施，直接關(guān)系到庫區(qū)人民的生命財(cái)產(chǎn)安全，一旦發(fā)生事故，后果將極為嚴(yán)重[1-3]。（剩余10114字）