基于蒙特卡罗法的多级黄土滑坡可靠性分析

滑坡稳定性分析目前工程中一般采用极限平衡算法。为了确定滑坡的稳定性系数，在计算过程中要将岩土体、计算模型简化，忽略一些影响滑坡稳定的次要因素，把影响滑坡稳定的各种主要因素作为确定参数。由于滑坡岩土体本身的非均质性及其参数确定具有随机性和变异性，用唯一的稳定系数不能客观地反映整个滑坡的真实安全程度。以陕西省夏呀河滑坡为例，通过野外详细调查、工程勘探和室内土工试验，初步确定该滑坡存在四级滑动面，难以用简单的单一滑动面计算其稳定性。因此，先采用推力传递系数法对滑坡体上的四级滑坡分别进行滑坡稳定性计算，得到该滑坡体上的四级滑坡稳定系数。再采用蒙特卡罗法对该四级滑坡进行可靠性分析，得到其可靠度，定量地表达夏呀河四级滑坡的安全程度。同时对比分析夏呀河四级滑坡的稳定系数和失稳概率，综合评价该滑坡体上的四级滑坡的稳定性及其风险概率，为滑坡的工程治理及预测预警、灾害危险性评价提供依据。

Reliability analysis of the multi-layer landslide in loess based on the Monte-Carlo simulating method

The slope stability analysis is the limit equilibrium method generally adopted in engineering designs. In order to calculate the stability factor of landslides, rock and soil and the calculating model will be usually simplified, some subordinate factors influencing landslide stability are neglected and all kids of primary influence factors of landslides will be commonly regarded as confirming parameters in calculation. Because of non-homogeneity of rocks and soils themselves of landslides and randomness and variability of confirming parameters, the real security degree of the entire landslide cannot be objectively reflected with the only stability factor. Take the Xiayahe landslide for example, it is confirmed by detailed field survey, geotechnical exploration and indoor geotechnical experiments that the landslide with four layers is difficult to calculate by a single slide layer. So stability factors of the four-layer landslide are firstly calculated with the overplus-pushing method. Then the probability and reliability of the four-layer landslide are obtained with the Monte-Carlo simulating method. Thus the security degree of the four-layer Xiayahe landslide is expressed quantitatively. Comparative analysis of the stability factors and instability probability of the four-layer Xiayahe landslide and comprehensive evaluation of the stability and risk probability of the four-layer landslide provide a basis for the engineering control and prediction and early-warning of the landslide and hazard danger evaluation.