基于MATLAB的BP神經(jīng)網(wǎng)絡(luò)在黃土液化評(píng)價(jià)中的應(yīng)用.pdf

1、學(xué)校代號(hào)；10731學(xué)號(hào)：P03舶140lol密級(jí)：公開(kāi)蘭州理工大學(xué)碩士學(xué)位論文Y910二76基于眥TLAB的BP神經(jīng)網(wǎng)絡(luò)在黃土液化評(píng)價(jià)中的應(yīng)用學(xué)位申請(qǐng)人姓名至曩筐導(dǎo)師姓名及職稱苤豎蔓副煎援培養(yǎng)單位查王捏!址專業(yè)名稱——蚩王捏論文提交日期!QQ笙!月12目論文答辯日期2四生縣!旦答辯委員會(huì)主席：生理重基于MAl[1AB的BP神經(jīng)網(wǎng)絡(luò)在黃土液化評(píng)價(jià)中的應(yīng)用AbstractThemalndistributiOnareaOf10essisal

2、sotheafeaofmacroselsm，which，intbehisIory，thebigearthquakehashappenedandinitiatedthemassiveloessliqueficationandlandslide，createdserjouslyeanhquakedisasterInreceⅡtyears，theiIlVestigationandnlelaboratorytesthaveindicatedma

3、tthe1iqueficationandthe1aIldslidecouldo咖edundertllehigllwatercontentand血ecertaillseismicactioⅡThisthesisin仃0ducedthemechanjsmofloess1iquefaction，themethodsofliquefactionjudgmcnt，tlleliquefactioⅡcoⅡditi0Ⅱaswellastheinnuen

4、tialfactorsItisdi伍cIlntoproposeaaccllratefo肋ulaofliquefactionbccausethefactorsoftheloessliquefactiona忿complex，mul矗pleandnonlinear髓eanificialneuralne時(shí)orksarecomposedofthemassivene盯oⅡswhicho唧izenon—linearauto—adaptedd”鋤ics

5、ystem，caIlpromulgatemenon—liIlearrclationswhichthedatasamplecOⅡtaiⅡsThosesyst鋤aIepTovidedwiththestrjveabjlityofstudy，association，fault—tolerallt卸dantijam，mayfacilitatemodelingtothemulti一鰣百nsarIdcomplexigIlo刪1ccfactoriⅡVa

6、ryingdegreeaⅡdleVelThespecialistswho擎aspmecomputerprogrammiⅡglanguagecanestablishBPnetworkmodel，whichdoesⅡotfaV0rthepromotionoft11eneⅣenetwOrktechⅡologyjnawayTheMAru出sof時(shí)arehasproVidedanavailableneⅣene時(shí)orktoolbox(NNI)。11

7、1isamcleu腳jzingtlleneⅣe那MoIktoolbox∞ND，conside血gthecharacteTisticof10essliquefication，hasestablished柳oBPne咐orkmodeleswhichdistiⅡctthepotemialanddamagede爭(zhēng)∞oft11eloessIique矗cation，aⅡdelaboratedthefcasibilityandtheusevalueo

8、fBPnetworkintheloess1iqueficationreferringtotheI’鋤gshanearthquakeaIldtheoverseasliquefiedmateriaByfhistbesis，wemaydrawt1)econdusjonmatitjseasyandeffectiVetoestablishBPne時(shí)orkmodelbyutilizingthenervene咖rkt001boxbasedonM越LA

9、BtoeValuatetheliqueficationoftlle10ess，whichneedn’tcaHyonthetediousprogmmmingEspecially，thismethOJdhastheverygoodversatilityandtheuseValueKeyWbrds：Loess；Liquefication；“queficationmechalljsm；“queficationjudgInent；MATLAB：N