基于matlab的bp神经网络在黄土液化评价中的应用-岩土工程专业论文.docxVIP

基于MAl[1．AB的BP神经网络在黄土液化评价中的应用A 基于MAl[1．AB的BP神经网络在黄土液化评价中的应用 A b s t r a c t The maln distributiOn area Of 10ess is also the afea of macroselsm，which，in tbe hisIory，the big earthquake has happened and initiated the massive loess liquefication and landslide，created serjously eanhquake disaster．In receⅡt years，the iIlVestigation and nle laboratory test have indicated mat the 1iquefication and the 1aIldslide could o咖ed under tlle higll water content and血e certaill seismic actioⅡ． This thesis in仃0duced the mechanjsm of loess 1iquefaction，the methods of liquefaction judgmcnt，tlle liquefactioⅡcoⅡditi0Ⅱas well the innuential factors．It is di伍cIln to propose accllrate fo肋ula of liquefaction bccause the factors of the loess liquefaction a忿complex，mul矗ple and non·linear．髓e anificial neural ne时orks are composed of the massive ne盯oⅡs which o唧ize non—linear auto—adapted d”锄ic system，caIl promulgate me non—liIlear rclations which the data sample cOⅡtaiⅡs．Those syst锄aIe pTovided with the strjve abjlity of study，association，fault—tolerallt卸d antijam， may facilitate modeling to the multi一鲥百ns arId complex igIlo删1cc factor iⅡ Varying degree aⅡd leVel． The specialists who擎asp me computer programmiⅡg language can establish BP network model，which doesⅡot faV0r the promotion of t11e neⅣe netwOrk techⅡology jn way．The MAru出sof时are has proVided available neⅣe ne时ork toolbox(NNI)。 111is amcle u脚jzing tlle neⅣe那MoIk toolbox∞ND，conside血g the characteTistic of 10ess liquefication，has established柳o BP ne咐ork modeles which distiⅡct the potemial and damage de争∞of t11e loess Iique矗cation，aⅡd elaborated the fcasibility and the use value of BP network in the loess 1iquefication referring to the．I’锄gshan earthquake aIld the overseas liquefied materia． By fhis tbesis，we may draw t1)e condusjon mat it js easy and effectiVe to establish BP ne时ork model by utilizing the ne咖rk t001box based M越LAB to eValuate the liquefication of tlle 10ess，which needn’t caHy the tedious progmmming． Especially，this methOJd h