A Prestack Seismic Inversion Method with Mixed-Driven Regularization

Abstract

The invention relates to a mixed-drive regularization pre-stack seismic inversion method, which belongs to the technical field of geophysical inversion. The method includes: 1) obtaining data and inputting an observation data set; 2) performing an unconstrained inversion on the above data to obtain an initial inversion result set; 3) calculating the spatial gradient of the initial inversion result in step 2); 4 ) classifying the spatial gradient in step 3) to obtain two types of spatial positions of the spatial gradient; 5) calculating the distribution characteristic function of the inversion result by the initial inversion result in step 2) and the position information in step 4); 6 ) Calculate the regularization term from the initial inversion result in step 2) and the distribution feature function in step 5); 7) Solve the objective function to obtain the final result. By using both the characteristics of the parameters to be inverted and the model assumptions of the inversion error to carry out regularization constraints on the inversion process, the stability of the inversion results is improved while the detailed characteristics of the inversion results are protected.

Description

technical field [0001] The invention relates to a mixed-drive regularization pre-stack seismic inversion method, which belongs to the technical field of geophysical inversion. Background technique [0002] Pre-stack seismic inversion is based on the relationship between seismic reflection amplitude and offset, and is an important technique for obtaining physical parameters of underground media. This technology can simultaneously obtain elastic parameters such as the compressional wave velocity, shear wave velocity and density of the underground medium, and has been widely used in the field of seismic exploration. The multi-parameter inversion of pre-stack seismic data is a highly nonlinear ill-posed problem, and a small change in the data will lead to a huge change in the solution; at the same time, coupled with the inevitable noise in the observation data, this further reduces the accuracy of the inversion results. stability. As a result, the detail changes in the inversi...

AI Technical Summary

Problems solved by technology

In the regularized inversion process, the smoothness of the parameters to be inverted is usually used as prior information to constrain the inversion results, but the traditional smoothness constraints will destroy the parameters to be inverted while improving the stability of the inversion results If you want to keep the detailed information of the inversion results, it may be difficult to effectively suppress the inversion noise

Images