Tunnel Least Squares Reverse Time Migration Imaging Method Based On Decoupled Elastic Wave Equation

China has become the country with the largest scale and difficulty in tunnel construction.In the process of tunnel construction,disasters such as water and mud inrush and landslides often occur.Using tunnel forward-prospecting technique to find out the geological conditions ahead of tunnel construction,and formulating reasonable treatment plans and construction plans in time is the current effective solution.In recent years,geophysical forward-prospecting technique have been widely used in tunnel construction.The seismic exploration has become an important method for forward-prospecting in tunnels due to its high interface imaging accuracy and long detection distance.In the seismic forward-prospecting of tunnels,high-precision migration imaging methods have always been a research hotspot.As the tunnel project continues to advance to western China,complex geological conditions have put forward higher requirements for the accuracy of seismic migration imaging.Hence,it is necessary to carry out systematic research on tunnel seismic migration imaging methods to improve the accuracy of tunnel seismic migration imaging.At present,some researchers have conducted preliminary studies on tunnel seismic migration imaging methods,but there are still some key issues:The tunnel seismic migration imaging methods are mainly ray-tracing migration and reverse time migration,and the resolution of the imaging result is limited.Moreover in the tunnel environment,P-wave and S-wave coupling crosstalk causes more low-frequency noise or artifacts,which seriously affect the migration imaging results.In view of the advantages of high resolution and amplitude balance,in this paper,we introduce least squares reverse time migration imaging method into tunnel seismic forward-prospecting.In order to further suppress the presence of low-frequency noise or artifacts in the migration imaging results caused by the coupled crosstalk of P and S waves,the tunnel least squares reverse time migration imaging method based on the decoupled elastic wave equation was further developed.The main work content and innovations are as follows:(1)Aiming at the accuracy and calculation efficiency of high-order staggered grid finite difference coefficients,a model-driven optimized difference coefficients calculation method and an artificial neural network accelerated calculation method are respectively proposed.The above methods effectively suppress the spatial numerical dispersion caused by the difference instead of the differential,and realize the high-precision forward modeling based on decoupled elastic wave equation;(2)Concentrate on the cumulative error problem caused by the mismatch of the least squares reverse time migration operator,we derive the adjoint wave equation of the decoupled elastic wave equation through matrix form,and construct the wavefield extrapolation operator pair and the migration imaging operator pair of least squares reverse time migration.Dot product test verifies the exact adjoint relationship of the operator pair,which ensures the stable convergence of the iterative algorithm;(3)On the basis of the high-precision forward modeling method and the exact adjoint operator pair,we adopt the wave impedance disturbance as the iterative convergence model parameter,and design a preconditioning operator based on pseudo-Hessian matrix to further improve the convergence of the iterative process.Finally,a tunnel least squares reverse time migration imaging method based on the decoupled elastic wave equation is formed;(4)In order to fully verify the applicability of decoupled elastic least squares reverse time migration for the complex geological environment in tunnels,we constructed lithological interface models,fault fracture zone models,karst cave models and complex models,carried out systematic numerical simulation migration imaging comparisons,and analyzed the imaging characteristics of migration imaging results.On this basis,field tests and applications were carried out to further verify the effectiveness,reliability and practicability of the method in this paper.