基于序列子空间法与MCMC算法的堤防冲刷风险分析  

Risk Analysis of Dike Erosion Based on Subset Method and MCMC Algorithm

作　　者：张文雅 郄志红[1] 吴鑫淼[1] ZHANG Wen-ya;QIE Zhi-hong;WU Xin-miao(College of Urban and Rural Construction,Agricultural University of Hebei,Baoding 071001,China;Tianjin Hydraulic Research Institute,Tianjin 300061,China)

机构地区：[1]河北农业大学城乡建设学院,河北保定071001 [2]天津市水利科学研究院,天津300061

出　　处：《数学的实践与认识》2023年第9期93-99,共7页Mathematics in Practice and Theory

基　　金：河北省水利科研与推广计划项目(2018-4)。

摘　　要：堤防的冲刷风险评价是堤防运行安全可靠性的重要手段.采用基于序列子空间法的MCMC算法对滹沱河石家庄市区段新建南堤进行冲刷风险分析,通过求解失效概率计算该堤防的冲刷结构可靠度,同时对比传统的蒙特卡罗方法(MCS);基于序列子空间法的MCMC算法计算的基础失效概率为3.061×10^(-5),MCS计算的基础失效概率为3.843×10^(-5).结果表明,两种方法计算的基础失效概率均在10^(-5)量级,并且在相同精度要求下基于序列子空间法的MCMC算法运行时间比MCS快了30倍,提高计算效率.The risk of embankment erosion is an important means to evaluate the safety and reliability of embankment operation.In this paper,the erosion risk of the new south embankment in Shijiazhuang section of Hutuo river is analyzed by MCMC algorithm based on sequence subset.The basic failure probability calculated by MCMC algorithm based on sequence subspace method is 3.061×10^(-5),and the basic failure probability calculated by MCS is 3.843×10^(-5).The results show that the basic failure probability calculated by the two methods is on the order of 10-5,and the running time of the MCMC algorithm based on the sequential subspace method is 30 times faster than MCS under the same accuracy requirements,which improves the computational efficiency.

关 键 词：风险分析 失效概率 序列子空间 MCMC算法 

分 类 号：TV871[水利工程—水利水电工程]