Simulation Research Of Geological Advanced Detection With DC Focus Method Based On Optical Fiber Sensing

Recently,DC focus method is used in advanced detection in TBM construction environment for its online exploration and easy installation.DC focus method based on fiber current sensor can increase the apparent resistivity observation accuracy of surrounding walls and distinguish the rock stratum conditions effectively.Besides,the method can obtain the real-time geological information of positions in front of and side of the tunnel face precisely.However,the method still has many difficult problems to break through both in theory research and engineering application.There are many research areas to be further studied like precision selection of fiber current sensor,studies of advanced detection distance and characteristic distinguish of water-bearing body.Furthermore,the method can provide reference for advanced detection of tunnels.In this study,based on the finite element method,simulation researches of three-dimensional forward models is carried out using secondary development with ANSYS parametric design language to deal with the above mentioned problems.Firstly,the validity of various schemes of finite element model is discussed to determine the specific process of steady-state electric field analysis with DC focus method based on fiber current sensor.The method is verified to have an orientated observation ability within a certain detection range by finite element simulation.Furthermore,the method can precisely distinguish the positions of geological and reduce the calculation errors of surrounding rock apparent resistivity.Secondly,models of observation errors of DC focus method based on fiber current sensor in propagation and synthesis process are analysed.The concrete influence of the fiber current sense on the measurment accuracy is clearly discussion.Therefore,the optimal measure accuracy of fiber current sense is determined as ± 5 mA after various simulations.Moreover,the criterion of water-bearing geological body existence is determined when the apparent resistance observation abnormal amplitude reaches 1.5 %.Then,the specific influences of other factors on detection distance are simulated.Simulation results shows that the volume of water-bearing body has large influence on advanced detection distance while the resistivity has a contrary effect.Meanwhile,the influence of metal iron track can often be ignored.For water-bearing body in front of the TBM whose volume is equal to 3 times of tunnels diameter,the advanced detection distance with DC focus method based on fiber current sensor can reach 2.75 times as long as the diameter of the tunnel.Finally,the prediction models based on annual neutral network is established to distinguish the characteristics of water-bearing body.The abnormal amplitudes after normalization are selected as the input parameters of the model.The model can not only distinguish the volume of the water-bearing body and distances deviating from the axis but also precisely predict the classifications of the directions.But,the prediction results of abnormal body volume is better than the distances deviating from the axis.