Coupling analysis of surrounding rocks in double-arch tunnel by FE and BP neural networks

A deformation prediction model of surrounding rocks in double-arch tunnel based on BP neural network was established with a practical project. Mechanical parameters of surrounding rocks were obtained by back analysis method based on measured data in left tunnel. Deformation prediction and stability analysis of surrounding rocks in right tunnel were carried out by finite element analysis. The proposed prediction model made feedback design on supporting parameters and construction method more effective. By Comparison, it is shown that the results of prediction are in good agreement with those measured in situ. It can meet the requirement of geotechnical engineering.