Drilling overflow prediction method, device, equipment and storage medium

By combining the RIME-ICEEMDA algorithm decomposition and denoising with an NGO-optimized CNN-SVM model, the problem of large overflow judgment error in existing technologies is solved, and high-accuracy prediction and control of drilling overflow is achieved.

CN122132934APending Publication Date: 2026-06-02CHINA UNIV OF PETROLEUM (BEIJING)
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
CHINA UNIV OF PETROLEUM (BEIJING)
Filing Date
2026-02-11
Publication Date
2026-06-02

AI Technical Summary

Technical Problem

Existing drilling overflow prediction schemes rely on engineers' experience, resulting in large overflow judgment errors, high false alarm rates, and difficulty in achieving accurate real-time monitoring and control.

Method used

The RIME-ICEEMDA algorithm is used to adaptively decompose and denoise the original overflow data, select the intrinsic mode components dominated by effective information, and optimize the CNN-SVM model for overflow prediction using the NGO algorithm, thereby improving data quality and model accuracy.

Benefits of technology

It improves the accuracy and generalization performance of drilling overflow prediction, reduces the false alarm rate, and achieves more accurate overflow detection and control.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122132934A_ABST
    Figure CN122132934A_ABST
Patent Text Reader

Abstract

This application discloses a drilling blowout prediction method, apparatus, equipment, and storage medium, relating to the field of drilling technology. The method includes: acquiring raw blowout data of the well to be predicted; decomposing the raw blowout data into intrinsic mode components at different time scales based on the RIME-ICEEMDA algorithm; selecting target intrinsic mode components dominated by effective information from all intrinsic mode components, and reconstructing the target intrinsic mode components to obtain denoised blowout data; inputting the denoised blowout data into a CNN-SVM model optimized by the NGO algorithm to obtain the blowout prediction result output by the CNN-SVM model.
Need to check novelty before this filing date? Find Prior Art