Pre-stack inversion method and system, storage medium and processor

By using a multi-source input method in prestack inversion, the two-dimensional profiles of prestack seismic data body, AVO attribute body and low-frequency model are input, which solves the problem of lateral discontinuity of the inversion profile in the prior art, and achieves higher lateral continuity and prediction accuracy.

CN120178326AActive Publication Date: 2025-06-20CHINA NAT PETROLEUM CORP
7 Cites 0 Cited by

Patent Information

Application Number
CN202311763549.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2023-12-20
Publication Date
2025-06-20
Estimated Expiration
2043-12-20

AI Technical Summary

Technical Problem

The existing pre-stack inversion method based on deep learning is prone to the transverse discontinuity of the inversion profile when the logging data is limited or the seismic data signal-to-noise ratio is low, and it is difficult to effectively improve the lateral continuity of the pre-stack inversion result.

Method used

The multi-source input method is adopted, including pre-stack seismic data body, AVO attribute body and low-frequency model two-dimensional profiles, and input it into the pre-trained two-dimensional network prediction model. By learning the low-frequency trend and geological characteristics of the target parameters, the lateral continuity of the inversion results is improved.

Benefits of technology

Through the multi-source input method, neural networks can better learn the low-frequency trends and geological characteristics of target parameters, significantly improving the lateral continuity of inversion results and the accuracy of prediction results.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The embodiment of the invention provides a pre-stack inversion method and system, a storage medium and a processor, and belongs to the technical field of oil-gas exploration and development. The method comprises the steps that a pre-stack seismic data volume, an AVO attribute volume and a low-frequency model of an inversion target area are acquired, and the low-frequency model is a low-frequency part extracted from a supervision model obtained based on multiple pieces of logging data of the inversion target area and a horizon interpolation method; respectively intercepting each two-dimensional profile of the pre-stack seismic data volume, the AVO attribute volume and the low-frequency model according to a predetermined rule, and inputting the two-dimensional profiles into a pre-trained two-dimensional network prediction model one by one to obtain a two-dimensional target parameter profile about the target parameter at the corresponding position of each two-dimensional profile in the inversion model; and synthesizing all the obtained two-dimensional target parameter profiles into an inversion model according to the position relation. According to the method, multi-source input is adopted, so that the neural network model can better learn the low-frequency trend and geologic body characteristics of the target parameters, and the transverse continuity of an inversion result is improved.
Need to check novelty before this filing date? Find Prior Art