Abstract
Aiming at the creep problem of Banshi Tunnel in Jilin province, the creep laws of rock are analyzed by the creep test, and the Cvsic model describing the creep characteristics of the tunnel is established. To obtain the creep parameters accurately, considering the advantages of Gauss process and differential evolution algorithm, coupling the two methods, a Gauss process‐differential evolution intelligent inversion method is proposed. According to on‐site monitoring data, the the creep parameters of the tunnel are accurately inverted. On this basis, the stability analysis of the tunnel and the selection of a reasonable construction plan are carried out. The research results show that to ensure the stability of the tunnel, the construction scheme of initial lining + pipe shed + advanced grouting anchor rod should be adopted. The research results have guiding significance for the long‐term stability evaluation of the tunnel.
Highlights
E back analysis of tunnel displacement by using field monitoring data provides a new idea for the study of creep parameters. e numerical back analysis method and the analytical back analysis method are the traditional methods to study creep parameters [16, 17]
Intelligent methods including Gaussian process (GP) and differential evolution (DE) algorithm have been widely used in geotechnical parameter inversion
Using the conjugate gradient method to obtain the optimal hyperparameters of GP, the conjugate gradient method is dependent on the initial value, falling into local minimum and iterations may not converge
Summary
2.1. e Problem of Creep Parameter Inversion. e inversion of creep parameters of tunnel surrounding rock is essential to optimize the parameters and find the optimal solution. E new generation of the population obtained by the selection operation is expressed as follows:. No as initia population e optimal nonlinear mapping between creep parameters and displacement is obtained. E combination of creep parameters is calculated by FLAC3D software, obtaining the displacement and establishing the learning sample. (6) e mapping relationship between creep parameters and displacement is established through Gaussian training process. E creep parameter to be randomly generated is the initial population, taking fitness function as evaluation index. If the fitness does not meet the requirements, mutation, crossover, selection, and other operations will be carried out until the maximum population iteration or fitness reaches the preset value
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
