摘要
综合考虑了参数分布类型、变异性及相关系数等因素,通过蒙特卡洛模拟开展了某均质边坡失效概率计算,探讨了三种因素对边坡失效概率的影响效应。结果表明,土体抗剪强度参数c(或)的变异系数与失效概率呈正相关性,的变异系数对失效概率的影响比c的变异系数更为显著。c、负相关时,失效概率随其相关系数增大而减小,反之相反。变异性系数对失效概率结果影响较相关系数更为明显。分别固定变异系数和相关系数时,不同分布类型得出的失效概率比较接近。但当c、分别服从伽玛分布和正态分布时,相应最大失效概率分别为15.50%和4.12%。这容易引起对边坡可靠性偏保守的估计。基于蒙特卡洛的可靠度分析可以定量描述各参数统计特性对边坡稳定性的影响,为边坡工程的风险评价提供参考。
Under a comprehensive consideration of the distribution types of parameters,the effects of the impacts from three factors on slope failure probability is discussed herein with the calculation made on the failure probability of a homogeneous slope through Monte Carlo simulation. The result shows that the variance coefficient of soil mass shear strength c（or ） is positively correlated with the failure probability,while the impact from the variance coefficient of is more significant than that from c. When c and are negatively correlated,the failure probability is decreased along with the increase of its correlative coefficient; and conversely it is opposite. The impact from the variance coefficient on the failure probability is more obvious than that from the correlative coefficient. When both the variance coefficient and the correlative coefficient are fixed respectively,the failure probabilities obtained from different distribution types are more closed. However,when c and are subject to the gamma distribution and normal distribution respectively,the corresponding failure probabilities are 15. 50% and 4. 12% respectively,which are prone to lead to more conservative estimations of reliability of the slope. Based on the Monte Carlo reliability analysis,the impacts from the statistical characteristics of all the parameters can be quantitatively described for providing references for the risk assessment of slope project.
出处
《水利水电技术》
CSCD
北大核心
2017年第12期195-198,206,共5页
Water Resources and Hydropower Engineering
基金
国家自然科学基金(51439003)
水利部公益性行业科研专项经费项目(201401029)
国家重点研发计划(2016YFC0402003)
关键词
边坡
蒙特卡洛法
失效概率
变异系数
相关系数
slope
the Monte Carlo method
failure probability
variance coefficient
correlation coefficient
