基于推广Bayes方法参数优化的滑坡稳定性评价

传统的极限平衡法在确定的计算工况下要求抗剪强度参数为常数,但由于各种参数的随机性、测试的误差、地质体的不均匀性、外界因素变化的随机性等特点,对滑坡稳定性评价起主要作用的抗剪强度参数从严格意义上来说均是随时间变化的随机变量,这就需要用破坏概率模型等非传统方法对滑坡进行稳定性评价。破坏概率模型的关键是滑坡抗剪强度参数的概率分布函数,但是在实际工程中,由于试验数据太少无法准确确定抗剪强度参数概率分布函数。本文提出了以区域统计规律得出的概率分布函数为先验分布函数,然后用推广Bayes方法确定该滑坡的概率分布函数,并以里沱河滑坡为例用破坏概率模型对其进行稳定性评价。

LANDSLIDE STABILITY ASSESSMENT BASED ON PARAMETERS OPTIMIZED BY PROMOTION BAYES METHOD

Limit equilibrium requires shearing strength parameters unchanged under a certain condition.However,as a result of each kind of parameter randomness,the test error,the geologic body nonuniformity,and the randomness of extra factor change,shearing strength parameters that play a main role on landslide stability are random variables which change with time,and the landslide stability assessment needs alternative methods such as destruction probabilistic model whose importantence is the probability distribution...