| This paper takes the slope of ion-type RE mine in south Jiangxi as an example,studies the application of landslide probability in slope stability analysis.The main research and conclusion of the thesis are as follows:(1)Landslide probability model and its application.General situation of the actual research according to the course of south Jiangxi ionic type rare-earth mine slope,the selected key influencing factors of landslide,through field investigation,geotechnical slope parameter measurement,data monitoring three ways for the 21 slope samples,using logistic regression models with deterministic coefficient CF landslide probability model is set up,by the landslide probability model and slope samples application instance.Selection of LongNan county DongJiang township foot hole test ore as a engineering example,a slope landslide probability model,including the classification and determine the subset of interval each subset sample interval uncertainty coefficient CF,slope parameters into deterministic logic regression coefficient CF values,the application of SPSS software,etc.,it is concluded that the landslide probability to solve the equation,and then forecast target probability of landslide,the probability apply the analysis results show that the landslide stability analysis.(2)Landslide probability model optimization.Landslide probability analysis of slope stability are defective its precision and accuracy,so the three kinds of optimization was carried out,including the state of slope landslide factors affecting subset selection range,value and reduce,the correlation among the influencing factors of landslide.Study chose two different subset selection classification method of subset interval optimization,the results show that the I class methods namely step taxonomy was suitable.Compare different slope state values for fitting analysis,the results show that the accuracy of the model is not affected by the values,but the results of the prediction deviation and precision are influenced by value,the larger the value,the higher the accuracy,but the possibility of the existence of deviation,the greater the practical application,should according to actual needs and slope sample selection accuracy slope state values.To reduce the correlation between influence factors can reduce landslide correlation between factors,can optimize the model,adopt the safety coefficient method,assuming a single factors for a particular variable to optimize landslide influence factors and the results show that the original paragraphs to return to the return of significance is not obvious in the optimization,to optimize the result obviously;Significance is very obvious and the original return item the return of the item,for its optimization,the optimization result is not obvious;Model of whole optimization effect is obvious.(3)Landslide probability and safety factor application comparison and relationship analysis.Respectively to solve the sample of the slope landslide probability and safety coefficient,and hierarchy,the results show that the landslide probability and safety coefficient of slope stability evaluation has good applicability and accuracy,and the project hierarchy is close to agreement.Function relation for the landslide probability and safety coefficient is derived,and the fitting and verification,can derive the function between the two,the result is expressed as: lnP/(1-P)= f(F_s)= 20.2-16.0 F_s. |
PDF Full Text Request