Real-Time Monitor Method of Soil Slope Stability in Seasonal Frozen Area

Abstract

Soil slope in seasonal frozen area subject to rainfall or freezing and thawing cycles, the shear strength of soil damage and lead to landslide hazard. In order to predict the landslide hazard through monitoring the soil slope stability real-time, the soil slope stability evaluation system is established. The Neural Network Algorithm can simulate elastic-plastic finite element method well, and using Genetic Algorithm to search cohesion(c) and internal friction angel(φ) which meet the accuracy requirements. And the intelligent parameter inversion model of soil shear strength (c, φ) based on displacement is established. So, the slope stability can be evaluated real –time based on displacement monitor. Application of this method in actual project shows that this method is accurate and effective.