ABSTRACT
The residual shear strength of cohesive soils is an important index in study and designing of geotechnic engineering projects, a neural network model that has been developed to estimate the residual shear strength with the physical property parameters of cohesive soils. It has been studied with the model that relation of the physical property parameters, which present the properties and conditions of cohesive soils, and its residual shear strength, based on the further analysis of residual shear strength mechanism and its influence factors. The results of calculation and analysis show that the model is more precise than the empirical formulas are. The better self-learning, and calculation capacity and mistake tolerance make the neural network models more exact and useful.
