论文部分内容阅读
目前考虑土体参数空间变异性的边坡可靠度分析所研究的边坡几何尺寸相对较小。为有效地分析考虑参数空间变异性的几何尺寸相对较大的边坡可靠度问题,提出了基于多重响应面与子集模拟的边坡可靠度分析方法。以一个坡高为24 m的两层非均质黏土边坡为例验证了提出方法的有效性,并探讨了正态、对数正态、极值I型、Gamma和Beta这5种概率分布类型对边坡可靠度的影响。结果表明,提出方法具有以下优势:1可以有效地计算考虑多参数空间变异性的低概率水平边坡可靠度;2可以较好地解决几何尺寸相对较大的边坡可靠度问题;3有较高的参数敏感性分析计算效率,可为调查概率分布类型和波动范围等参数统计特征对边坡可靠度的影响提供技术支持。此外,参数概率分布类型对边坡可靠度具有重要的影响,常用的正态和对数正态分布分别用于表征参数概率分布特征时,可能会分别高估和低估边坡失效概率。
At present, considering the spatial variability of soil parameters, the slope geometry studied is relatively small. In order to effectively analyze the reliability of slope with relatively large geometrical dimensions considering the spatial variability of parameters, a slope reliability analysis method based on multiple response surface and subset simulation is proposed. Taking a two-layer heterogeneous clay slope with a height of 24 m as an example, the validity of the proposed method is verified and five probability distributions of normal, log normal, extreme I, Gamma and Beta are discussed Influence of Types on Slope Reliability. The results show that the proposed method has the following advantages: (1) It can effectively calculate the reliability of low-probability horizontal slopes considering the spatial variability of multi-parameters; (2) Solve the reliability problem of slopes with relatively large geometrical dimensions; (3) The high parameter sensitivity analysis and calculation efficiency can provide technical support for investigating the influence of statistical characteristics such as probability distribution type and fluctuation range on slope reliability. In addition, the type of parameter probability distribution has an important influence on the slope reliability. When the commonly used normal and lognormal distributions are respectively used to characterize the probability distribution of parameters, the probability of slope failure may be overestimated and underestimated.