Fuzzy Neural Reliability Analysis Of Soil Slope Stability Under Rainfall Conditions

In recent years, the application and study about analysis of slope stability study inthe method of reliability analysis is developing rapidly. The application of reliabilitytheory of slope engineering experienced the transition by the random reliability to thefuzzy reliability. At present, the best method to fully consider the fuzzy random anduncertainty question on the slope is combined with the fuzzy mathematics theory toanalyze the slope fuzzy random reliability analysis. But, on the whole, the slope of thefuzzy random reliability analysis method is still at the starting stage. The parameters offuzzy number and safety coefficient of the membership functions still have manysubjective factors, and need the accumulation of experiences of engineering practice.The paper contains the main work and results are as follows:①The first, analysed the effects of rainfall on the soil parameter from the angle ofdeformation and failure of slope, rainfall infiltration factor and early warning andforecast of rainfall type landslide, and then established the relationship model based onthe influence of rainfall on soil parameters. On the establishment of relations modelbetween effective rainfall and soil parameters, derivated the formula of effective rainfalland unit weight of soil (γ). Using Gauss formula in fit the test data of the cohesion Cand saturation.②The secend, using fuzzy logic and neural network coupling methods to constructa kind of slope hazard analysis model based on fuzzy random reliability analysis; andfurther discussed the model of program steps based on Matlab numerical software, andthe Anfis system generation.Combined with previous research experience and ownconclusion, and choose a more reasonable form to express to parameter fuzzymembership function.On the basis of this, combined with the rainfall on the reliability analysis of theeffects of parameter model, established under the condition of rainfall slope stabilityreliability of intelligent evaluation system, and further discussed based on Matlabnumerical software model of Anfis system program, and generates the system analysisof the interface.③Finally, use the constructed model of the slope risk evaluation fuzzy reliabilityand study mine waste dump slope in the south of Chongqing Zhong Liangshan as anexample,and the fuzzy reliability index of the parameter sensitivity analysis.The results show that, the model method can be carried out slope risk analysis and evaluation thatunder the conditions of rainfall.The systems analysis results of engineering examplesshow that the system has ability of intelligent diagnosis and judgement. Theconclusion can be used for reference in practical engineering.