Inverse Scattering Inversion Algorithm Based On The Second-order Born Approximation

Inversion has been one of the key targets for exploration geophysical and applied mathematics research.Although the full waveform inversion has the advantages of high precision and high resolution.The selection of the initial iterative velocity model will seriously affect the result of the full waveform inversion.so it is important that a more accurate velocity model is provided as the initial iterative model of the full waveform inversion.On the other hand,the inverse scattering inversion algorithm under high frequency approximation has higher accuracy and fast calculation characteristics,and its calculation result can be used as the initial iterative model of full waveform inversion.Based on the above reasons,this paper proposes a new class of inverse scattering inversion methods.For the complexity of the medium,the constant velocity background inverse scattering inversion operator and the variable speed background inverse scattering inversion operator are discussed respectively.In order to further investigate the effectiveness and superiority of the proposed method,the paper numerically simulates the nonlinear inverse scattering inversion operator for the three-dimensional medium model,and compares the experimental results with the linear inverse scattering inversion operator results.The numerical results show that the nonlinear inverse scattering inversion operator has a smaller error.The inverse scattering inversion algorithm based on the second-order Born approximation can overcome the weak perturbation limitation of the inverse scattering inversion algorithm based on the first-order Born approximation and can adapt to strong disturbance Inversion of velocity media.On this basis,this paper also proposes an inverse scattering inversion-like heterogeneous parallel algorithm based on the domestic chip "Shenwei 26010".In the three-dimensional case,the numerical results show that the heterogeneous parallel algorithm in the background of constant velocity and variable speed the computational efficiency can reach 0.88 and 0.58,respectively.The parallel efficiency is basically stable with the increase of the number of nodes.The heterogeneous parallel algorithm of nonlinear inverse scattering inversion has strong scalability.