Page 122 - 《水资源与水工程学报》2025年第1期
P. 122

!36 "!1 #                         & ' ( ) & * + , -                               Vol.36No.1
               2025 $ 2 %               JournalofWaterResources&WaterEngineering                 Feb.,2025

            DOI:10.11705/j.issn.1672-643X.2025.01.13


            Ôãä匵¦W BiTCN-Attention-LSSVMfOMN


                                       1,2,3        1,3         1,3        1,3       1,2,3
                                †‡ˆ        , q‰V , |yŠ , ,‹Œ ,  Ž
                  (1. ¼F—õa0 Zb0G, ¼F Á¤ 650201;2. E9a0 Zþ…á9!?@pqr, ¡Á FU 210098;
                                  3. ¼Ff4–ºZbCDá✦CDAB4•, ¼F Á¤ 650201)
               > ?: ¯§4Æp€0iº¦ÿ„Ï@íêÏcֈ{©ªô©éÞstžª:?WÑ{ª³, –#¯}
               .%ƒœÔÏ@øÌµŸ¶·(BiTCN)、 ډ'p´(Attention) žz–_I¤¡Ôp(LSSVM) {è,<½Jðñ
               ¿/iº。uiºb• BiTCNX¨、 P–f€Ô©*Ï@íê4{ô©éÞst, £ç Attention p´‘xiº'Ú
               ƒcðñ’s{sîª:Â7, À° BiTCN-Attention æ~ÿ„P{Â7Àç LSSVMiº4ÆÇðñ, zPÖ 2 f
               ‹Ò{íêþµ¶¯iº{ðñ{«。ò µ¶S¤: c LSSVM、CNN-LSSVMž TCN-LSSVM’6,BiTCN-At
               tention-LSSVMiºðñ{Á(³´ü†3z», ¦è,<ñ&œZ¦ðñ490#¢é{iºß~ží·Ñ;
               BiTCNc Attention {’“®/ȯ¢Ý4–+Ï@íê4{’“éÞst, ° BiTCN-Attention –+{Â7Àç
               LSSVM4ÆÇðñ›*t[Ý{ðñÑÈ, íêþÃ^ÿ„Pˆ{–é¯iº{0È'。
               1@A: è,<ñ&œZ¦; ½Jðñ; œÔÏ@øÌµŸ¶·; ډ'p´; z–_I¤¡Ôp
               BC<DE:TV698.11   FGHIJ:A    FKLE:1672643X(2025)01011811

                              BiTCN-Attention-LSSVM Modelingforseepage
                                          predictionofearth -rockdams

                                                                                     1,3
                                                    1,3
                                                                   1,3
                       FUShuyan  1,2,3 ,YANGShiyong ,CHENDehui ,WANGZixuan ,OUBin              1,2,3
                (1.CollegeofWaterConservancy,YunnanAgriculturalUniversity,Kunming650201,China;2.NationalKeyLaboratoryof
                   WaterDisasterPrevention,HohaiUniversity,Nanjing210098,China;3.ResearchCenterforSmartManagementand
                   MaintenanceofSmallandMediumSizedWaterConservancyProjectsinYunnanProvince ,Kunming650201,China)
               Abstract:Conventionalmachinelearningmodelsoftenfailthetaskofidentifyinglongterm dependencies
               andlocalimportancewhendealingwithtimeseriesdata.Toaddressthisproblem ,weproposedacoupled
               modelforseepagepredictionofearthandrockdamsbasedonbidirectionaltemporalconvolutionalnetwork
               (BiTCN),attentionmechanism (Attention),andleastsquaressupportvectormachine(LSSVM).The
               modelusesBiTCNtocapturethelongterm dependenciesinthetimeseriesdatafrom bothforwardand
               backwarddirections ,introducestheattentionmechanismtohelpthemodelfocusonthekeylocalfeatures
               relatedtotheprediction,andinputsthedeeplyprocessedBiTCN-AttentionfeaturesintotheLSSVMmodel
               forbetterpredictionresults.Toanalyzethepredictioneffect,themodelwasusedtoprocesstwodifferent
               datasets.TheresultsofcaseanalysisshowthatcomparedwithLSSVM,CNN-LSSVMandTCN-LSSVM,
               theevaluationindexesofBiTCN-Attention-LSSVM modelarethebest,andtheproposedmodelshows
               higheraccuracyandstabilityinthepredictionofpiezometerwaterlevelintheearth-rockdam.Thecombi
               nationofBiTCNandAttentioncanbetterextracttheinterdependencerelationshipintimeseriesdata.Input
               tingthefeaturesextractedbyBiTCN-AttentionintoLSSVM forpredictioncanachievebetterprediction
               performance.Afterthedatasetisexpanded ,thelearningabilityofthemodeliseffectivelyimproved.
               Keywords:piezometerwaterlevelinearth-rockdam;seepageprediction;bidirectionaltemporalconvolution

                !"#$:20240726; %&#$:20241031
                '()*: !"wx#0%»(3(52069029、52369026);“ }¨}Ç” Zc›¡¢89#$%»]x(3(2023490411);
                        ¼Ff—õ%&ABo/'(r¬(3(202401BD070001071)
                +,-.: •.((1976—), Æ, ¼F:/N, OP, QR, ÈPSTU, AB€ÔCD'0cZß89œ„。
                qr+,: A ƒ(1983—), K, ±FÁ¯N, OP, QR, ÈPSTU, AB€ÔZC®X89b¾c½¾0-。
   117   118   119   120   121   122   123   124   125   126   127