基于LHS-MC的堤防渗透破坏风险分析  

Seepage Failure Probability Analysis of Embankments Based on LHS-MC Method

作　　者：雷鹏[1,2,3] 陈晓伟[2,3] 张贵金[3] 田大作[4] 

机构地区：[1]南京水利科学研究院水文水资源与水利工程科学国家重点实验室,江苏南京210098 [2]水利部堤防安全与病害防治工程技术研究中心,河南郑州450003 [3]长沙理工大学水利学院,湖南长沙410114 [4]湖南省水利水电科学研究所,湖南长沙410076

出　　处：《人民黄河》2014年第10期45-47,共3页Yellow River

基　　金：国家自然科学基金资助项目(51409023);水文水资源与水利工程科学国家重点实验室开放研究基金资助项目(2011491611);水利部堤防安全与病害防治工程技术研究中心开放研究基金资助项目(2013006);湖南省教育厅优秀青年基金资助项目(12B008)

摘　　要：以有限元法分析为基础,结合拉丁超立方抽样和蒙特卡罗方法(LHS-MC),可对不同洪水位下的堤防渗透破坏风险进行分析。在分析中,考虑堤防土体渗透系数的不确定性,将其视为服从正态概率分布的随机变量,计算得到堤防相应的最大水力坡降,并通过统计N次计算结果,根据渗透破坏功能函数,计算堤防在相应水位下的破坏概率。首先,采用拉丁超立方抽样方法对渗透系数进行N次抽样;然后,对每一次抽样进行一次确定性有限元渗流不同水位条件下的渗透破坏概率计算。结果表明:堤防渗透破坏的概率随着水位的上升而变大,渗透系数的随机性也对堤防渗透破坏影响较大,拉丁超立方抽样和蒙特卡罗方法应用于水工结构破坏风险分析,可有效提高计算效率。Based on the stochastic finite element method,the probability analysis had been done on the seepage failure of embankments by means of LHS-MC method. Considering the uncertainty of the soil permeability as random variables which obeyed some kinds of probability normal distri-bution,the stochastic seepage of the embankment was analyzed by the suggested method. Firstly,the multiple permeability samples were genera-ted;secondly,the maximum hydraulic gradient for each sample could be gained by the certainty finite element analysis,comparing to the permissi-ble hydraulic gradient and according to the seepage failure function,seepage failure probability of the embankment could be calculated in different conditions. The case study shows that the failure probability increases with the increasing of water level and the randomness of permeability signifi-cantly affects the seepage failure . The results show the effectiveness,practicality and reliability of the suggested method.

关 键 词：堤防 渗透破坏 蒙特卡罗方法 拉丁超立方抽样 

分 类 号：TV871[水利工程—水利水电工程]