Page 167 - 《水资源与水工程学报》2022年第3期
P. 167

!33 " ! 3 #                       & ' ( ) & * + , -                               Vol.33No.3
               2022 $ 6 %               JournalofWaterResources&WaterEngineering                 Jun.,2022

            DOI:10.11705/j.issn.1672-643X.2022.03.21


                    '1 IGA-BPNN2¬RbŽû¿ðe


                                           1,2        1,2         3       3        1,2
                                    r"Á , D* , Š¦Ã , Š Ä , C Å
                   (1. ;÷k6 WZlmV—*¥ß=>&'?@A, ;÷ 300350;2. ;÷k6 ‚.lm6n, ;÷ 300350;
                                3. Râ=eßW¶B)§û8V ÁLW‚)¶à, 3¬ {±P 617000)

                 D E: s?¿W¶Ù಍XÃYÕËޙÙï>ïÈ, ›ñ/ÃYFŽ*ÃYê¡Óª¶•‡E, Ÿ\è
                 #³uW¶Ùà²ÃYê¡s@, Í Aè†71 IGA-BPNNW¶Ùà²ÃYê¡ïÈde。OPZ[ IGA
                 v BPNNó"þ&ГñИ™Æ×, ÷)ì IGA·wõ™、 ßåø)–', á;~ BPNNÏìëì!õX
                 ÃY¥ªïÈÐ。IeÈ?@œ, È'ÃY¥ªïÈÐ{kµv/­ý$~ 11%;IGA-BPNNde*ÄÂ
                 deµ», ®ïȖ"“((¶•‰ó§AàÍ<, Z„ïȉ¾§w, žsÄÂÃeW¶ÙÃY(^Í
                 În‹。
                 FGH: #³uW¶Ù; à²ÃYê¡ïÈ; å×úò7¾;BP bÑÏì; ÃYê¡
                 8IBJK:TV312   LMNOP:A    LQRK:1672643X(2022)03016306

              VibrationresponsepredictionofahydropowerstationbasedonIGA-BPNN

                                                                                3
                                                                                               1,2
                                                                     3
                                      1,2
                                                       1,2
                        CHENJinquan ,WANGHaijun ,LIUJuanli,LIUTong,LIANGChao
                  (1.StateKeyLaboratoryofHydraulicEngineeringSimulationandSafety,TianjianUniversity,Tianjin300350,China;
                       2.SchoolofCivilEngineering,TianjianUniversity,Tianjin300350,China;3.ErtanHydropowerPlant,
                                YalongRiverHydropowerDevelopmentCo.,Ltd.,Panzhihua617000,China)
                 Abstract:Inordertoachieveintelligentpredictionoftheoveralloperationconditionsofthehydropower
                 plantstructurevibrationandsolvetheproblemsofthecouplingofmultiplevibrationsourcesandrandom
                 nessofvibrationresponse ,weproposedapredictionmodelbasedonimprovedgeneticalgorithm-back
                 propagationneuralnetwork(IGA-BPNN)tostudythepowerhousevibrationresponseofariverbedhy
                 dropowerstation.Firstly ,theinitialweightvaluesandthresholdsofBPNNareoptimizedbyIGA,which
                 ischaracterizedbytheadvantagesofefficientparallelismandglobalsearch ,andthenthepredictedvalues
                 ofstructuralvibrationdisplacementareprocuredbytrainingtheBPNNnetwork.Theprototypeobserva
                 tionexampleshowsthatthemaximumrelativeerrorofthepredictedvibrationdisplacementsatthemeas
                 urementpointsdoesnotexceed11%;theIGA-BPNNmodelissignificantlysuperiortoothermodelsin
                 termsofpredictionaccuracyandconvergenceperformance ,indicatingthatthepredictionmethodiseffec
                 tiveandfeasible.Thisstudycanprovideareferenceforvibrationresearchofothertypesofhydropower
                 plants.
                 Keywords:riverbedhydropowerstation;plantvibrationresponseprediction;improvedgeneticalgorithm
                 (IGA);backpropagationneuralnetwork(BPNN);vibrationresponse
                                                               ¶  [1-2] 。çsW*¶ñ<X, W¶Ùà²
            1 ./01
                                                               ¥ß‡EèU…(^0&'29vp                    [3] 。B+
                 W¶Ù಍XÃY‡E–½Ý, ®ªÙ、 ¶                       ¿‹ÃYÈ4žŸ–ˆÿK>à²ÃYÙï, £
            ´“W‚/ÃYO6@, W;)¶ªQ“಍X                            ۟È5Å·µ, ìížZ[,§È'ÃYÐ@
            Q5ÃY€œ¿ ýˆ‰¶、 >‚¶“JY                             v಍XÃYꡘ™ïÈ, UŠ?¿W¶Ùà

                !"#$:20210923; %&#$:20220301
                '()*: =>&'()±â.3(2016YFC0401905); ;÷g45/678.3(18JCYBJC22300)
                +,-.: ®8k(1997-), a, \j¥»K, ÓM(^P, (^‰FsWZW¶lm。
                /0+,: h¾¢(1978-), a, =]GàK, LM, NO, (^‰FsWZW¶lm。
   162   163   164   165   166   167   168   169   170   171   172