Evaluation of Tunneling-Induced Lateral Pile Response by an Artificial Intelligence Optimization Algorithm

Abstract

The efficient assessment of tunneling effects on adjacent existing piles is significant for underground constructions. In this study, a new analytical approach was developed to rapidly assess lateral pile responses due to tunneling. A nonlinear Pasternak foundation model considering the unloading effect (NPFM-U), which took into account the nonlinearity of the pile–soil interaction and the attenuation of the soil resistance caused by tunneling, is proposed. Inspired by the minimum potential energy principle, an accurate and efficient artificial intelligence optimization algorithm––chaos radial movement optimization (CRO)––was developed to optimize the minimum value of the total potential energy of the tunnel–soil–pile system in which the proposed NPFM-U was utilized, and to obtain the lateral responses of the piles. The reliability of the proposed method was subsequently validated by comparing it with data from centrifuge and field tests. Parametric studies on the influence of the parameters of the CRO algorithm’s maximum iteration number, number of particle groups, tunnel diameter, pile modulus, pile diameter, pile length, soil undrained shear strength, reduction factor, and pile–soil horizontal distance were also performed.