Rock stratum drilling in-situ identification method based on physical information machine learning and application thereof
By collecting multi-source signal parameters during drilling in cold regions and utilizing physical information machine learning methods based on CEEMD and LSTM networks, real-time identification of rock strata and rock mass strength in frozen formations in cold regions was achieved. This solved the problem of insufficient lithology identification during drilling in traditional methods, and improved mining efficiency and safety.
Patent Information
- Application Number
- CN202511919708.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-18
- Publication Date
- 2026-03-20
AI Technical Summary
In the process of drilling and blasting excavation in open-pit mines in cold regions, traditional rock mass property assessment methods cannot achieve in-situ identification during the drilling process, resulting in lag in blasting parameter design and insufficient judgment of slope stability, which affects mining efficiency and safety.
A physical information machine learning-based in-situ rock strata identification method is adopted. By collecting multi-source signal parameters during the drilling process, such as drilling pressure, drill rod torque, drill rod rotation speed, drilling speed, and drilling vibration and acoustic emission signals, dynamic signal features are extracted using CEEMDAN and VMD decomposition techniques. The physical inversion equation is constructed by combining LSTM network and Businesk elastic half-space theory to achieve real-time identification of rock strata strength and rock mass type.
It enables accurate identification of rock strata type and rock mass strength during drilling in frozen strata in cold regions, supports intelligent drilling control, optimizes drilling parameters, and improves mining efficiency and safety.