A method and device for generating geological interpretation conclusions of acoustic wave remote detection imaging images

By using hybrid discriminator generative adversarial network (HDGAN) and reinforcement learning, a geological interpretation library of acoustic long-range imaging images is constructed to generate accurate geological interpretation conclusions. This solves the problem of inaccurate interpretation conclusions in existing technologies and achieves fast and reliable geological interpretation.

CN122260486APending Publication Date: 2026-06-23CHINA NAT PETROLEUM CORP +1
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
CHINA NAT PETROLEUM CORP
Filing Date
2024-12-20
Publication Date
2026-06-23

AI Technical Summary

Technical Problem

The geological interpretation of imaging results from existing acoustic remote detection imaging logging technology is subject to multiple interpretations, relies on the professional skills of the interpreters, and is time-consuming, leading to inaccurate interpretation conclusions.

Method used

A hybrid discriminator generative adversarial network (HDGAN) combined with reinforcement learning is used to construct a "forward modeling-imaging-interpretation conclusion" library for long-range acoustic detection of geological anomalies. Text reports are generated by a generator, and single-sentence and single-word discriminators are used for discrimination and gradient strategy updates to form fast and accurate geological interpretation conclusions.

Benefits of technology

It enables rapid and accurate qualitative interpretation of acoustic long-range imaging images, reduces interference from human factors, improves the reliability and consistency of interpretation conclusions, avoids underfitting problems, and enhances the ability to interpret complex geological structures.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure FT_1
    Figure FT_1
  • Figure FT_2
    Figure FT_2
  • Figure FT_3
    Figure FT_3
Patent Text Reader

Abstract

The application discloses a kind of acoustic wave far detection imaging image geological interpretation conclusion generation method and device, specifically a kind of machine learning method using hybrid discriminator to generate joint reinforcement learning of generative adversarial network, first, construct acoustic wave far detection geological anomaly body " forward modeling-imaging-interpretation conclusion " library;Second, far detection imaging image is input generator to generate text report;Then, using single sentence discriminator discriminates the generated text, using single character discriminator discriminates the generated text;Finally, using reinforcement learning rewards the model, finally forms the interpretation conclusion corresponding to the acoustic wave far detection image.The application realizes the fast generation of acoustic wave far detection well logging imaging image geological interpretation report, reduces the interpretation deviation problem caused by human factors, greatly improves the reliability and uniformity of geological interpretation conclusion.
Need to check novelty before this filing date? Find Prior Art