Comprehensive monitoring of talus slope deformation and displacement back analysis of mechanical parameters based on back-propagation neural network

Landslides are regarded as significant geological hazards across the world, causing serious economic losses and casualties. The understanding on deformation characteristics and failure mechanisms of landslides plays the vital roles in slope stability evaluation and reinforcement design. In this study, the deformation characteristics and failure mechanism of the Xiaomiaoling talus slope were analyzed based on field monitoring data. In addition, as it was difficult to measure the shear strength parameters of the rock–soil mixture due to its complex spatial structure and variable material composition, a displacement back analysis based on the back-propagation neural network (DBA-BPNN) was proposed to determine the shear strength parameters of the rock–soil mixture. The analytical results show that deformation of the Xiaomiaoling talus slope was that of a typical traction landslide, which has the characteristics of progressive failure, and major slope deformation was triggered by excavation and rainfall. According to field monitoring data, the shear strength parameters of the rock–soil mixture could be determined. The predicted cohesion and internal friction angle of the rock–soil mixture were 10.84 kPa and 19.51°, respectively, and the predicted and test values were in good agreement. The method proposed in this paper can provide references for the design and construction in geotechnical engineering.