
基于Copula函数的边坡可靠度高效分析方法
方旭, 冯晓波
基于Copula函数的边坡可靠度高效分析方法
Efficient Ways to Analyze Slope Reliability Based on Copula Function
将Copula函数与高效蒙特卡洛方法结合,提出了含相关随机变量的边坡可靠度高效分析方法。以两个岩质边坡稳定性问题为例验证了所提方法的有效性。结果表明,该方法相比于直接蒙特卡洛方法在保证边坡失效概率的准确性的同时计算效率更高。Copula函数可以构造出具有不同相关结构的岩土体参数的联合概率分布,与高效蒙特卡洛方法(即蒙特卡洛重要抽样方法和子集模拟)结合能高效地处理含多种相关随机变量的边坡可靠度计算问题,相比于现行的Nataf变换方法结果更能体现岩土边坡真实稳定性。此外,该方法也能高效地计算含有复杂的隐式功能函数的边坡可靠度,研究成果拓展了高效蒙特卡洛方法在边坡可靠度分析中的应用。
In this paper, the copula function is combined with efficient Monte Carlo Method, methods which can efficiently analyze the slope reliability problem involving correlated random variables are proposed. Two rock slope stabilities are studied to verify the effectiveness of the proposed methods as examples. The results show that these methods are more efficient than the direct Monte Carlo Method and the accuracy of the results is guaranteed. Based on the copula function which can be used to construct the joint distribution of different related structural parameters of rock and soil mass, efficient Monte Carlo Method (Monte Carlo important sampling method and subset simulation) can efficiently deal with the slope reliability problem with correlated random variables. Comparing with the existing Nataf transformation method, the results can reflect real geotechnical slope stability more. In addition, the methods can efficiently analyze the reliability of slope with complex implicit functions. The results expand the application of Monte Carlo important sampling method and subset simulation in slope reliability analysis.
Copula函数 / 边坡可靠度 / 相关随机变量 / 高效蒙特卡洛方法 {{custom_keyword}} /
copula function / slope stability / correlation random variables / efficient Monte Carlo Method {{custom_keyword}} /
表1 边坡滑动面强度参数的统计特征Tab.1 Statistical characteristics of strength parameters of sliding surface of slope |
项目 | f 1 | f 2 | c 1/kPa | c 2/kPa | α 1 | α 2 |
---|---|---|---|---|---|---|
均值 | 0.939 | 0.277 | 249.09 | 39.22 | 30° | 59°08′ |
标准差 | 0.123 | 0.010 | 84.34 | 4.90 | 7°15′ | 6°20′ |
分布类型 | 对数正态 | 对数正态 | 对数正态 | 对数正态 | 正态 | 正态 |
表2 不同可靠度计算方法的边坡可靠度结果Tab.2 Results using different slope reliability methods |
可靠度计算方法 | N | Copula函数 | |||||||
---|---|---|---|---|---|---|---|---|---|
Guassian | Frank | Plackett | No.16 | ||||||
Pf | 误差/% | Pf | 误差/% | Pf | 误差/% | Pf | 误差/% | ||
直接蒙特卡洛(MCS) | 1×106 | 0.027 4 | 0 | 0.028 3 | 0 | 0.028 7 | 0 | 0.032 6 | 0 |
重要性抽样(ISM) | 1×104 | 0.022 0 | 19.71 | 0.028 6 | 1.09 | 0.024 3 | 15.32 | 0.030 0 | 7.98 |
子集模拟(SS) | 2×102 | 0.027 2 | 0.73 | 0.029 5 | 4.24 | 0.029 5 | 2.79 | 0.034 4 | 5.52 |
表3 左岸边坡滑动面强度参数的统计特征Tab.3 Statistical characteristics of the strength parameters of the sliding surface of the left bank slope |
项目 | f 1 | f 2 | c 1/kPa | c 2/kPa |
---|---|---|---|---|
均值 | 0.400 0 | 0.450 0 | 30.0 | 50 |
标准差 | 0.600 0 | 0.076 5 | 5.4 | 10 |
分布类型 | 对数正态 | 对数正态 | 对数正态 | 对数正态 |
表4 不同方法计算的可靠度结果(天然工况)Table.4 Results of different slope reliability methods (Natural condition) |
可靠度计算方法 | N | Copula函数 | |||||||
---|---|---|---|---|---|---|---|---|---|
Guassian | Frank | Plackett | No.16 | ||||||
Pf | 误差/% | Pf | 误差/% | Pf | 误差/% | Pf | 误差/% | ||
直接蒙特卡洛(MCS) | 1×107 | 0.000 220 | 0 | 0.000 389 | 0 | 0.000 397 | 0 | 0.000 476 | 0 |
重要性抽样(ISM) | 1×105 | 0.000 292 | 24.65 | 0.000 355 | 8.74 | 0.000 345 | 13.10 | 0.000 389 | 18.27 |
子集模拟(SS) | 2×103 | 0.000 232 | 5.45 | 0.000 358 | 7.97 | 0.000 343 | 13.60 | 0.000 488 | 2.52 |
表5 不同方法计算的可靠度结果(降雨工况)Table.5 Results of different slope reliability methods (Rainfall condition) |
可靠度计算方法 | N | Copula函数 | |||||||
---|---|---|---|---|---|---|---|---|---|
Guassian | Frank | Plackett | No.16 | ||||||
Pf | 误差/% | Pf | 误差/% | Pf | 误差/% | Pf | 误差/% | ||
直接蒙特卡洛(MCS) | 1×106 | 0.001 4 | 0 | 0.001 4 | 0 | 0.001 3 | 0 | 0.003 5 | 0 |
重要性抽样(ISM) | 1×104 | 0.001 2 | 14.3 | 0.001 9 | 35.71 | 0.001 8 | 38.46 | 0.002 5 | 28.57 |
子集模拟(SS) | 8×102 | 0.001 2 | 14.3 | 0.001 7 | 21.43 | 0.001 3 | 0 | 0.002 9 | 17.14 |
表6 不同方法计算的可靠度结果(地震工况)Tab.6 Results of different slope reliability methods (Earthquake condition) |
可靠度计算方法 | N | Copula函数 | |||||||
---|---|---|---|---|---|---|---|---|---|
Guassian | Frank | Plackett | No.16 | ||||||
Pf | 误差/% | Pf | 误差/% | Pf | 误差/% | Pf | 误差/% | ||
直接蒙特卡洛(MCS) | 1×107 | 0.271 7 | 0 | 0.268 5 | 0 | 0.268 5 | 0 | 0.264 6 | 0 |
重要性抽样(ISM) | 1×105 | 0.299 5 | 10.23 | 0.296 7 | 10.50 | 0.296 7 | 10.50 | 0.291 0 | 9.97 |
子集模拟(SS) | 2×103 | 0.305 6 | 12.48 | 0.308 3 | 14.82 | 0.308 3 | 14.82 | 0.305 3 | 22.41 |
1 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
2 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
3 |
陈祖煜,陈立宏,王玉杰,等. 滑坡和建筑物抗滑稳定 分析中的可靠度分析和分项系数设计方法[C]//高坝 大库安全建设与风险管理高端论坛,2011.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
4 |
王禹玺,李典庆,唐小松.边坡系统可靠度区间估计方法[J].武汉大学学报(工学版),2019,52(11):941-950.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
5 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
6 |
唐小松,李典庆,周创兵,等. 不完备概率信息条件下边坡可靠度分析方法[J]. 岩土工程学报,2013,35(6):1 027-1 034.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
7 |
范明桥,盛金保. 土强度指标φ, c的互相关性[J].岩土工程学报,1997(4):100-104.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
8 |
LOW B K. Reliability analysis of rock slopes involving correlated nonnormals [J]. International Journal of Rock Mechanics and Mining Sciences, 2007,44(6).
{{custom_citation.content}}
{{custom_citation.annotation}}
|
9 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
10 |
唐小松. 基于Copula理论的岩土体参数不确定性建模与可靠度分析[D]. 武汉:武汉大学,2014.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
11 |
张曼,唐小松,李典庆.含相关非正态变量边坡可靠度分析的子集模拟方法[J].武汉大学学报(工学版),2012,45(1):41-45,53.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
12 |
曹子君.子集模拟在边坡可靠性分析中的应用[D]. 成都:西南交通大学,2009.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
13 |
张明.结构可靠度分析:方法与程序= Structural Reliability Analysis: Methods and Procedures[M]. 北京:科学出版社,2009.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
14 |
李洪双,马远卓. 结构可靠性分析与随机优化设计的统一方法[M]. 北京:国防工业出版社,2015.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
15 |
祝玉学. 边坡可靠性分析[M]. 北京:冶金工业出版社,1993.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
{{custom_ref.label}} |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
/
〈 |
|
〉 |