基于Monte-Carlo的滑坡失稳概率计算研究

Study on the Probability of Landslide Failure by Monte-Carlo Method

普通的滑坡稳定性计算,只能对滑坡的安全性做一个直观的计算,而滑坡的失稳概率却无从得知。利用蒙特卡洛（Monte-Carlo）法可确定滑坡的失稳概率,但传统应用蒙特卡洛（Monte-Carlo）计算滑坡失稳概率时,常认为c和φ值服从正态分布,其实在不同情况下c和φ值服从多种分布。基于此,探讨了岩土体参数c和φ服从不同分布形式时的滑坡失稳概率计算。结果表明,不考虑c和φ值的变异系数和偏度系数计算得出的失稳概率比考虑c和φ值的变异系数和偏度系数计算出的失稳概率偏高,且2种失稳概率误差为10%～25%。上述研究对滑坡失稳概率计算有一定的参考价值。

Landslide is the most common geological disaster with the most severe damages. By ordinary stability calculation we can only obtain the stability,and by using Monte-Carlo method we can determine the failure probability of landslide,in which case,however,the values of c and φ are regarded as to obey normal distribution. In fact,the values of c and φ obey different distributions under different circumstances. In view of this,we discuss the calculation of failure probability with different c and φ values in different distributions. Results reveal that the calculated failure probability in consideration of variation coefficient and skewness coefficient is smaller than that not considering variation coefficient and skewness coefficient. And the calculation error under these two circumstances ranges from 10% to 25%. By this research,the accuracy of failure probability calculation is enhanced.