| The stability evaluation of the slopes has been the a hot area of geotechnical engineering.At present,there are two main methods of the stability evaluation of the slopes,including safety coefficient method and reliability method.However,both methods exist some disadvantages:the former can not consider the uncertainty and reliability of the geotechnical parameters and the dynamic load imposed on the slopes,which make the safety level of the slope engineering indistinctly and can not evaluate the slopes clearly;the latter,though it can take the reliability of various factors,is not consummate in the elements of the correlation of geotechnical parameters,the statistics of geotechnical parameters under the condition of a small sample and the evaluation indictors of dynamic reliability,which result in the bad practicality and accuracy.In view of the above several problems,this paper puts forward relative solutions,and calculates programs of the reliability analysis of slopes.Principal results are as follows:(1)Give an outline of the fundamental theory of the stability analysis and reliability analysis of slopes,and discuss the advantages and disadvantages of different reliability methods;put forward the searching method of the probabilistic critical slip surface based on the response surface method,which Using quadratic polynomial response surface function instead of difficult to seek partial derivatives of the implicit expression of safety coefficient to obtain the minimum reliability index and the corresponding sliding surface;by changing the parameters of slope stability analysis method and the variation coefficient of correlation coefficient and the obedience of probability distribution,the static reliability sensitivity analysis of the slope stability.(2)Analyze three data in the actual project by A-D test,discovering that normal distribution probability distribution is not always the distribution which c and?do not always obey,Logarithmic normal distribution,extreme value distribution and Weibull distribution might be the distribution that c and ? obey.It will lead to a larger error using the normal distribution to simulate the probability distribution directly.(3)To describe the correlation between parameters of introducing Copula theory,using the joint probability distribution function to establish Copula parameters and reliability analysis of the slope,found that negative correlation between shear strength parameters and has great influence on the reliability of the results,the results do not consider the correlation of slope reliability analysis is conservative,bring unnecessary waste.(4)The slope under the condition of small sample data reliability analysis,statistical analysis of the random weighting method is introduced for the first time the geotechnical parameters,variation simulation of different alternative Copula function AIC value and slope failure probability,and then get the confidence interval of slope failure probability using percentile method.A simple example and an engineering example show that the method can solve the reliability analysis under the condition of small sample data.(5)The permanent displacement of slopes as the slope stability evaluation index of dynamic reliability,considering the stochastic earthquake by Newmark on the finite displacement theory of permanent displacement,and the permanent displacement of the cumulative probability distribution of any possible permanent deformation corresponding to the failure probability,make engineering personnel according to slope the importance of the slope and the repair degree of difficulty to choose different slope allowable deformation as control standard,the overall coordination of security and economic contradictions.(6)Considering the randomness of its seismic intensity,through seismic site hazard analysis to get the probability distribution of intensity,then the probability of total probability rule to obtain any permanent displacement of the corresponding slope;considering the uncertainty of geotechnical parameters and correlation on the effects of the different factors such as variation of parameters and the parameters of the probability distribution of dynamic reliability. |