嵌巖長樁承載特性的人工神經(jīng)網(wǎng)絡(luò)模型的建立與研究

1、河海大學(xué)碩士學(xué)位論文嵌巖長樁承載特性的人工神經(jīng)網(wǎng)絡(luò)模型的建立與研究姓名：康小兵申請學(xué)位級別：碩士專業(yè)：港口、海岸及近海工程指導(dǎo)教師：魯子愛20050301AbstractNow，allarenotveryintensivetOthepressure，distortionanddestructivemechanismofthelongrock—socketedpilesAndsystemictestingresearchonthemare

2、notenoughLearningfromthemaininfluencingfactorsofpilebearingcapacityandsettlement，weknowthatthesoilaroundpileisimportanttopilebearingcapacityanditssettlement，besidesthepileintensityanditsgeometryBecausethecharacteristicof

3、soilisverycomplex，itconducestOthatestimatingpilebearingcapacityaccuratelyisverydifficultSo，empiricalandsemi—empiricalmethodsareoftenappliedtopileultimatebearingcapacityandsettlementButallofthesehavedifferencewiththefactd

4、uetotheassumptionfalledshortofitOwingtotheANNhasitssuperiorityandadaptabilitywhendealingwithempirical，non—linearandcomplexquestion，theauthorwanttoestablishitsmodelanduseittOforecastpileultimatebearingcapacityandsettlemen

5、tInthepaper，theauthorcollectedthedataofthelongrock—socketedpilesinHangzhouItreducedinputdataservingtheANNutilizingPrincipalComponentAnalysisTheauthorhasestablishedtheANNmodelaboutrock—socketedpilesinHangzhouandhasimprove

6、donit(especiallycombinedtheGA)IthasconfirmedthefeasibilityandCOlTeCtflessofthemodelthroughcomparingtheresultsAnalyzingontheinfluencingfactorsoftheultimatebeatingcapacityandsettlement，ithasdrawnsomeconclusionwhichavailtoe