The Research Of The First-arrival Travel Time Tomography Based On Least Square Method

The velocity of seismic wave propagate in subsurface media has always been the object of seismic exploration.By inverting the velocity of the underground structure,the depth,size,and specific location of the underground anomaly can be shown.Because the lack of resources,all walks of life have a lot of demand for resources,and velocity inversion can provide effective help for the exploration and collection of the oil and gas resources.The first-arrival travel time tomography method is a kind of seismic method that uses the first arrival of the earthquake to invert the velocity of the subsurface media.This technology has the advantages of the fast process,accurate results and so on.By comparing the received first arrival travel time information with the travel time in the established forward model,an inversion calculation is performed to obtain the slowness of the underground medium.After many times iteration,the velocity of the subsurface media can be better shown.In this paper,the forward modeling simulation and inversion of the first-arrival travel time tomography and the optimization of the entire imaging process are studied.This paper describes the basic principles of first-arrival travel time tomography,and optimizes the algorithm according to the different problems that need to be solved.The application of the least squares method to the first-arrival travel time tomography is studied.When the condition number is large,the damping parameter is added to the basic formula of the classic least squares to make the inversion process of the matrix stable.And when the travel time data exists some defects,in this case,the first arrival time data of seismic wave can be reweighted to improve the resolution of the inversion result.In the forward modeling of first-arrival travel time tomography,this paper studies the traditional shortest path method and the finite difference method,and applies the fast deduction to solve the first-time travel time of seismic waves.When the first arrival time of the seismic wave is obtained by the FMM,and then the trajectory of the seismic wave is inversely determined from the detection point according to the already-acquired market.The cost of calculation in the forward modeling can be reduced,and the real first arrival wave timing layer can be improved.Analyze the efficiency of imaging,moreover,the method's calculation process conforms to the laws of physics,and the operation process is unconditionally stable.In the inversion process of first-arrival travel time tomography,the inverse of the solution matrix will be complicated in the past.In order to simplify the calculation process,the optimization process needs to be performed.In this paper,we introduce the relation of steepest descent method,Gauss-Newton's method and conjugate gradient method.Through the comparison of the tomographic results obtained by the three methods with the specific model,it is shown that the conjugate gradient method has the best effect,the highest resolution,and fewer iterations in the entire process.