Search

The stochastic medium technique (SMT) combined with genetic programming (GP) was used to predict ground subsidence due to tunnelling. SMT was employed to generate theoretical models and the parameters in the models were optimised using GP, which was trained using tunnelling engineering data. The approach was applied to the estimation of ground movement at two projects in China, the Hanjiazhuang section of the Bao-Fu highway tunnel and the Taipingliang section of the Zhang-Shi highway tunnel. The results obtained using the SMT-GP method were in good agreement with field measurements and provided the smallest error when compared with traditional finite element methods., The stochastic medium technique (SMT) combined with genetic programming (GP) was used to predict ground subsidence due to tunnelling. SMT was employed to generate theoretical models and the parameters in the models were optimised using GP, which was trained using tunnelling engineering data. The approach was applied to the estimation of ground movement at two projects in China, the Hanjiazhuang section of the Bao-Fu highway tunnel and the Taipingliang section of the Zhang-Shi highway tunnel. The results obtained using the SMT-GP method were in good agreement with field measurements and provided the smallest error when compared with traditional finite element methods.