Study On The Characteristics Of Rain Infiltration And Its Influence On Slope Stability

Landslide is a main geological hazard and one of the important geological and engineering problems that have been under research over the world. Because factors and mechanism are variable and complex, it's difficult to predict it. Penetrating hydraulic pressure and flow field distributing of slope are important factors that affect the stability and transformation. Moreover, rainfall is frequently derivational factor. In order to analyze the influence of rainfall on the unstaturated soil slopes' stability, on the basis of the saturated-unsaturated seepage theory, the extending method of traditional limit equilibrium method, ANN (Artificial Neural Network), the thesis have been studied in details about the stability analysis, condition forecast of the landslide and obtain some significative conclusion.Explaining the fundamentals of the steady-unsteady saturated seepage and its initial and boundary conditions. On the base of basic seepage physics equation, regarding pressure hydraulic head as a basic unknown quantity to deduce saturated-unsaturated seepage differential equation, the FEM format of differential equation is deduced based on Galerkin method of weighted residual. Numerical models were used to study how the infiltration into a slope varies with the rainfall intensity, rain fall duration, ante-rainfall, initial pore water pressure and soil properties and how the infiltration influences the stability of a slope with seep program.Regarding rainfall seeping as an important factor of soil slope stability analysis, Analyzing the change of saturated-unsaturated seepage field for soil slope, the distribution of pore water pressure is calculated and introduced into the formula of safety factor to find the law under different parameter. It can be evaluated for the stability of the slope, and this offer the decision the reliable theoretical basis.A global optimal algorithm is put forward based on alternative and iterative algorithm to optimize the structure of the neural network The predicting model to estimate stability of the slope are built. This method make the neural network will have super nonlinear mapping ability and generative ability. Numerical example illustrates that the newlyintroduced method based the artificial neural network can provide more accurate analyzing results. It is a new effectual method to forecast slope stability.In this paper, the influence of main factor on the seepage field and stability is discussed, which is very helpful for appropriate design of soil slope, effective protection and forecast of landslide.