Lithology identification methods, devices, storage media and electronic equipment

CN118468074BActive Publication Date: 2026-06-30CHINA PETROLEUM & CHEMICAL CORP +1
View PDF 6 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
CHINA PETROLEUM & CHEMICAL CORP
Filing Date
2023-02-08
Publication Date
2026-06-30

AI Technical Summary

Technical Problem

Existing lithology identification technologies lack real-time and adaptive capabilities, making it impossible to perform analysis and prediction while drilling. This makes it difficult to optimize drilling parameters during oil and gas drilling, increasing drilling risks.

Method used

By acquiring the mechanical drilling parameters uploaded by the logging-while-drilling device, and after preprocessing and normalization, a multi-layer neural network model is used for lithology prediction, including a first logging prediction model, a second logging prediction model, and a lithology classification model, to identify lithology categories in real time.

Benefits of technology

It achieves real-time and adaptive lithology identification, enabling early identification of continuous rock strata boundaries, optimizing drilling operations, reducing the risk of complex drilling failures, increasing drilling speed, and reducing drilling time.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118468074B_ABST
    Figure CN118468074B_ABST
Patent Text Reader

Abstract

This invention provides a lithology identification method, apparatus, storage medium, and electronic device. The method includes acquiring mechanical drilling parameters uploaded by a logging-while-drilling (LOD) device, calling a trained lithology prediction model, preprocessing the LOD parameters to obtain input data, and processing the input data using the lithology prediction model to obtain a predicted lithology category based on the drilling depth. In this method, the LOD device is placed several meters above the drill bit, so the mechanical drilling parameters are the earliest available data. Based on this, this application uses mechanical drilling parameters for lithology classification, which can quickly and accurately classify the lithology of the current drilling by continuously adapting to changing downhole conditions. This allows the boundaries between continuous rock layers to be identified as early as possible, alleviating the technical problem of poor real-time performance in current lithology identification technologies.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Lithology identification method based on machine learning

    CN114462471A

  • Rock sample intelligent identification method and system based on image deep learning

    CN114581666A

  • Adaptive meta-learning lithology identification method and device based on attention

    CN114708517A

  • Shale gas reservoir lithology intelligent identification method based on electric imaging logging image

    CN115019144A

  • Oil well reservoir property detection method and device

    CN110284873A