Research On Safety Prediction Of Composite Soil Nailing Wall Based On Elman Neural Network

Composite soil nailing walls are widely used in underground engineering ofhigh-rise building and subway project. The safety of composite soil nailing walls areinfluenced by multi-factors, it is difficult to predict its security using the traditionalmethod that first establish the mechanical model, and then use the constitutive relationof deformation. Against the nonlinear uncertainty characteristics which is possessedbetween safety performance and its influencing factors of composite soil nailing walls.This article will use the artificial neural network which has a strong nonlinear mappingability and a high degree of fault tolerance to make a prediction.Through analyzing the basic structure of the composite soil nailing walls and thefundamental role of the various components, this paper makes a systematic andcomprehensive analysis about disconnect the water curtain, soil nail wall, prestressedbolt down drainage, earthworks excavation, and establish a index system which affectsthe safety of composite soil nailing walls. On this basis, it will use the Elman neuralnetwork to build the safety prediction model of composite soil nailing wall. Throughconcrete examples, the effect of the practical application of the model is verified and theElman model prediction is compared with the BP neural network model prediction, it isproved that the model is feasible and effective, and this model has a higher predictionaccuracy.