ABSTRACT
2D analysis and application of the convergence-confinement method are commonly used to evaluate rock-support interaction. However, this approach reaches its limits when the ground exhibits large deformation and/or when the support is very stiff and installed close to the tunnel face. In that case, 3D numerical simulations are required. An alternative based on a machine learning approach is proposed here, in order to obtain a reliable evaluation of the displacements at equilibrium state. Using data previously obtained from numerical simulations, a simplified Symbolic Regression (SR) approach is proposed to evaluate its applicability for obtaining a reliable evaluation of the displacements at equilibrium. The results show that this model performs well with the small data set used in this study and can be considered as a useful alternative in the engineering field.
