Study On The Law Of Surface Subsidence And Settlement Prediction Of Double-line Shield Tunnel

Shield construction method is now becoming the most commonly used subway tunnel construction methods,because of the advantages of safety,fast,high degree of automation and not disturbing the ground transportation and facilities.And two-line parallel tunnel is the most widely used subway line design form yet.The study of surface subsidence and the prediction of surface subsidence caused by the construction of shield tunnel in the two-lane subway has significant meaning of reference and guidance for subway construction.The main research work in this paper includes the following aspects:(1)First of all,by referring to the literature,the construction principle of shield method and the formation mechanism of surface subsidence are described in detail.Then,based on the double-line superimposed Peck formula,C = L / i is proposed(L represents the tunnel axis spacing,i represents the surface subsidence groove width coefficient),and the change of the shape of the surface settlement of the double-line shield tunnel with the C value is discussed: As C slowly increases the shape of the curve gradually change from the V-type to W-type,and then transformed into two V-type.The results shows that C = 2 and C = 7 are the critical points of two situations which are "single peak" turning into "double peaks" and "double peaks" turning into "single peak".(2)A three-dimensional finite element model of two-line shield tunneling was established by using MIDAS / GTS NX finite element software,and the results were compared with the measured data.The comparision results basicly matchs,and verifies the reliability of numerical simulation.And the law of land subsidence of two-line shield tunnel construction is summed up.The influence of different influencing factors on the surface subsidence of double-line shield tunnel is discussed.Then the influence of tunnel axis distance,soil property parameters and asynchronous excavation distance on surface subsidence is obtained.(3)Using the wavelet neural network method to predict the cumulative settlement of the surface and comparing with the actual monitoring data,it is founded that the method of usingthe wavelet neural network to predict the deformation data is feasible and can reduce the monitoring to a certain extent error.However,in the actual monitoring process,due to the actual situation is more complex,the cumulative surface settlement caused by external conditions will affect the training of wavelet neural network in certain extent,so in the case of small sample monitoring data,using the wavelet neural network method to predict surface subsidence is lack of good stability,and the predicted data can provide reference for practical engineering,but it can not replace the actual monitoring work.