| With the increasingly close relationship between human activities and environment,various engineering activities will inevitably encounter slope stability problems,which has been a hot research issue in geotechnical engineering.There are many uncertainties in the geotechnical parameters of actual slope engineering.In the risk assessment of slope stability,these uncertainties are not taken into account,and the parameters are mostly calculated with fixed values,ignoring their variability.Importance measurement analysis can simultaneously consider changes in geotechnical parameters and allow them to change within their entire distribution range(i.e.uncertainty range).According to the importance index of each parameter,relatively important and unimportant parameters can be distinguished,the relative contribution of the uncertainty of each parameter to the model output uncertainty,or the interaction between different parameters can be evaluated.In the slope engineering activities,focusing on the uncertainty of the parameters with high importance index can greatly reduce the uncertainty of the output response,thereby effectively improving the efficiency of slope engineering design and optimization.In this study,the importance analysis method is used to analyze the influence of the uncertainty of each parameter on the slope stability.The main research contents and conclusions are as follows:1?An improved QGA algorithm combined with the limit equilibrium method is proposed to search the critical slip surface of slope.Firstly,a dynamic adjustment strategy is used to update the quantum rotation gate,and the limit equilibrium method is used to calculate the fitness(i.e.the safety factor of slope,Fs).Secondly,the penalty function is applied to punish the solutions that do not meet the constraint conditions of slip surface,rather than discarding them directly in the search process,which can reduce their probability of being selected into the next generation.The proposed method has high computational efficiency and global optimization ability in the analysis of nonlinear,high-dimensional and multimodal problems.Finally,the results of four slope engineering cases are consistent with the existing literature,which verifies the accuracy of this method.Compared with other optimization methods,the results obtained by the proposed method are less uncertain and have good calculation efficiency,so they can provide strong support for optimization problems in slope engineering.2?The ability of three copula functions to construct bivariate distribution of shear strength parameters in the analysis of three-dimensional reliability of seismic slopes is first studied.The specific steps of three-dimensional reliability analysis of seismic slopes with copula function sampling method are given.Then,the local sensitivity analysis of two seismic slope cases are performed and compared with the results of Monte Carlo(MC)method,which verifies the accuracy of the proposed method.Finally,the effects of shear strength parameters,horizontal seismic amplification factors,seismic wave periods,amplification factor and Kendall rank correlation coefficient which characterize the negative correlations between shear strength parameters on failure probability are investigated.The results show that increasing the shear strength parameters and their negative correlation can reduce the failure probability,and increase the seismic parameters will lead to the increase of failure probability,which is not conducive to slope stability;in the first stage of seismic wave period,the failure probability presents obvious waveform characteristics.3?Based on the Sobol sequence and Least angle regression(LARS)algorithm,a global sensitivity analysis method for slope stability is proposed,and the specific steps of the calculation of the least angle regression algorithm are given.In order to verify the proposed method,the importance measurement indexes of the parameters was calculated by MC method based on variance.The results of the proposed method are in good agreement with those of MC method.In the three slope engineering case studies,the influence of each parameter on the safety factor(Fs)of slope was analyzed,and the importance ranking of the influence of each parameter on the safety factor was obtained.In addition,the influence of the negative correlation between the shear strength parameters on the importance index of each variable is discussed.The results show that when the shear strength parameters are independent or related,the importance indexes and ranking of each variable are different.For comparison of calculation efficiency,in the process of the global sensitivity analysis of LARS method,the operation number of the function of Fs is far less than that of MC method.The former is only 1/(n·N+1)of the MC method,where n is the number of random variables and N is the samplesize of the variable.The above results show that the computational efficiency of the proposed method is significantly higher than that of MC method.4?Research shows that the global sensitivity analysis based on variance will inevitably bring about loss of parameter information.The moment-independent importance index satisfies"global,quantifiable,universal and moment-independence",which can better reflect the cumulative effect of input variables on the output response.This chapter combines the moment-independent importance analysis method with Kernel density estimation(KDE)and Orthogonal polynomial estimation(OPE)to analyze the cumulative effect of different parameters on the safety factor of slope.The numerical results are consistent with the results of the traditional variance-based MC method,which verifies the accuracy of the proposed method.In addition,the influence of the negative correlation between the shear strength parameters on the importance index of each variable is discussed.The results show that when the shear strength parameters are independent or correlated,the importance indexes and ranking of each variable are different,which should not be neglected.In the calculation process,the Sobol low deviation sequence and the principle of equal probability translation are used to simulate the geotechnical parameters,the sample size required is smaller than that of the ordinary MC method.In addition,when the moment-independent method is used to calculate the unconditional output response and the conditional output response,the opration number of the function of Fs are only(n+1)/(n·N+1)of MC method.Therefore,the proposed method provides an efficient calculation method for the analysis of the influence of the multi factor importance of slope stability.The results of the three slope engineering cases show that the influence of geotechnical parameters on slope stability is different.5?Based on the foregoing research,a failure probability importance analysis method was developed.The variance importance analysis method based on LARS and the moment-independent importance analysis method based on KDE and OPE were used to study the cumulative effects of different parameters on the failure probability.The results are consistent with the traditional variance-based MC method,which verifies the accuracy of the proposed method.The results of the three engineering case studies show that the correlation between the shear strength parameters will greatly affect the failure probability,resulting in a large difference in the importance index and ranking of each variable.Moreover,there is a significant difference between the importance ranking of Fs and the importance ranking of the failure probability p_f,which means that Fs cannot be used as the only criterion in the risk assessment of slope stability.At the same time,in the risk assessment based on the failure probability,the variables with larger importance index values should be considered to reduce their uncertainty. |
PDF Full Text Request