A Data-Driven Deep Machine Learning Approach for Tunnel Deformation Risk Assessment

Abstract

The shallow overburden pipe jacking over operatio n tunnel construction project in chalk stratum has the risk of defo rmation of the soil layer and the existing tunnel, which increases t he difficulty of pipe jacking over construction, and the risk assess ment and control become the key technology for the safe and succ essful completion of the construction. Aiming at the problems of t he current deformation risk assessment and control method, such as the assessment system is not comprehensive, systematic and ob jective enough, the prediction accuracy is not efficient enough, an d there is a lack of quantitative analysis, etc., a deformation risk a ssessment and control method is proposed to combine the heuristi c optimization algorithm of human behaviour and deep machine l earning algorithm for pipe jacking up to and across operation tun nels on shallow overburden of chalky sand stratum. Firstly, by an alyzing the construction process of pipe jacking tunnel, the defor mation risk factors of the construction process and the deformati on risk control scheme are given; then, a deformation risk assess ment and control algorithm with improved deep limit learning m achine is proposed by combining human heuristic optimization al gorithm; finally, the proposed deformation assessment and contr ol model is applied to the deformation risk assessment and contro l problem of pipe jacking over operation tunnel on shallow overb urden of pulverised sand stratum, and a finite element computati onal model is used to construct the data. Finally, the proposed def ormation assessment and control model is applied to the problem of deformation risk assessment and control in a tunnel with shallo w overburden in chalky sand stratum by using finite element com putational model to construct the data set, training the deformati on risk assessment and control model, and using the monitoring d ata as the test set to validate the validity of the proposed model al gorithm, and solving the problem of the poor prediction accuracy of the control algorithm for deformation risk assessment and con trol of a tunnel with shallow overburden in a tunnel with shallow overburden in chalky sand stratum.