| With the rapid development of transportation construction in China,the number of highway tunnels has been increasing year by year,and at the same time,they are also facing the problem of tunnel construction causing changes in the tunnel itself and surrounding environment.Although the traditional mining method is often used in mountain tunnel construction,it is easy to have safety accidents because its section is prone to overcut and undercut.This study relies on Qinhuangdao Jiaoshan Tunnel project,on the basis of existing research at home and abroad,Through investigation and research on physical geography,geomorphology,engineering geology and hydrology of Qinhuangdao area,combined with measured data,based on the random medium theory,analyze the surface displacement and deformation caused by gas production tool blasting tunnel excavation.At the same time,finite element simulation is used to study the displacement and deformation laws of the strata in the tunnel under static load and gas production equipment blasting load.Finally,a prediction model for the peak particle velocity(PPV)of gas production tool blasting and the resulting surface subsidence was proposed.The specific research work carried out includes:(1)Based on random medium theory and MATLAB software,write a back analysis program to fit the measured data of surface subsidence,Inversion of section convergence areaΔA and stratum influence Angle β,study the relationship between the two and the relationship betweenΔA and tunnel depth Z.When the tunnel burial depth is about 20.53 m,the convergence area.ΔA.approaches 0.AsΔA gradually increases,tanβ gradually decreases.When the burial depth of the tunnel is about 20.53 m,tan β is about 0.447.(2)Firstly,the static load model of Jiaoshan tunnel is established by MIDAS GTS / NX,and the surface settlement and stress state of the model were extracted to analyze the stratum displacement and deformation of the tunnel under static load.At the same time,establish a simulation model for the blasting load of gas production tools,and analyze the displacement deformation of the strata and the propagation law of vibration velocity in the rock caused by gas production tool blasting construction.(3)Through analysis of measured data,it was found that the accuracy of PPV prediction was improved by 3.14 times when considering frequency influence compared to not considering frequency influence.At the same time,the strength of the relationship between the detonation center distance,maximum single shot explosive quantity,and frequency with PPV was analyzed.Among them,frequency has the greatest impact on PPV prediction,with a relationship strength of 0.95225.(4)Based on rolling prediction,Back Propagation(BP)neural network was established to predict future surface deformation of tunnel section DM-1.The standard particle swarm optimization(PSO)algorithm is improved for its slow convergence and easy to fall into local optimization,a gas production tool blasting PPV prediction model(IPSO-BP)considering frequency effects is proposed by taking the distance between detonation centers,maximum single shot explosive quantity,and frequency as input parameters of the network,and PPV as output parameter. |
