Knowledge-Based Machine learning for Real-Time rock strength testing while Drilling: Bridging Simulation and Reality
文献类型：期刊论文
| 作者 | Bai, Jun3; Wang, Sheng3; Liu, Liu2; Xu, Zhengxuan1; Li, Shaojun2; Chen, Minghao1; Luo, Zhongbin3; Li, Bingle3; Hou, Jin1 |
| 刊名 | MEASUREMENT
 |
| 出版日期 | 2025-03-31 |
| 卷号 | 246页码:17 |
| 关键词 | Intelligent Real-time In-situ Rock Strength Testing Physics-Informed Machine Learning Explainable AI Methods Multi-Scenario Adaptability |
| ISSN号 | 0263-2241 |
| DOI | 10.1016/j.measurement.2025.116664 |
| 英文摘要 | This paper proposes a method for predicting rock strength based on the fusion of physical information from while-drilling tests. Using Boussinesq's elastic half-space theory, a fundamental mechanical model for uniaxial compressive strength based on drilling parameters is developed. Through model experiments, we derive empirical formulas for the uniaxial compressive strength of carbonaceous slate, granite, and sandstone. Meanwhile, we construct a sample library for different lithological types using field test data.The findings indicate that the mechanical model performs poorly in anisotropic formations such as carbonaceous slate. While fully datadriven AI methods show high dependency on the quantity of labeled data, supervised learning models (KNN, RF, GBDT, ANN, 1D-CNN) can achieve high accuracy given sufficient labeled data. However, unsupervised learning techniques like K-means clustering exhibit limited effectiveness. The fusion of physical principles with machine learning techniques addresses these challenges effectively, achieving high prediction accuracy even in the absence of labeled data, with the best model achieving an R2 of 0.82. Additionally, SHAP interpretability methods are employed to explore the influence of drilling parameters on the model's decision-making and their interactions. This framework combines the interpretability of physical models with the adaptability of AI, ensuring effective generalization from simulated experiments to real geological environments. |
| 资助项目 | National Natural Science of China[42072339] ; National Natural Science of China[44341702388] ; National Natural Science of China[U19A2097] ; State Key Laboratory of Geohazard Prevention and Geoenvironment Protection[SKLGP2022Z006] ; Everest Technology Research Proposal of Chengdu University of Technology[80000-2020ZF11411] |
| WOS研究方向 | Engineering ; Instruments & Instrumentation |
| 语种 | 英语 |
| WOS记录号 | WOS:001398207400001 |
| 出版者 | ELSEVIER SCI LTD |
| 源URL | [http://119.78.100.198/handle/2S6PX9GI/37253]  |
| 专题 | 中科院武汉岩土力学所 |
| 通讯作者 | Wang, Sheng; Liu, Liu |
| 作者单位 | 1.China Railway Eryuan Engn Grp Co Ltd, Chengdu 610031, Peoples R China 2.Chinese Acad Sci, Inst Rock & Soil Mech, State Key Lab Geomech & Geotech Engn, Wuhan 430071, Peoples R China 3.Chengdu Univ Technol, State Key Lab Geohazard Prevent & Geoenvironm Prot, Chengdu 610059, Peoples R China |
| 推荐引用方式 GB/T 7714 | Bai, Jun,Wang, Sheng,Liu, Liu,et al. Knowledge-Based Machine learning for Real-Time rock strength testing while Drilling: Bridging Simulation and Reality[J]. MEASUREMENT,2025,246:17. |
| APA | Bai, Jun.,Wang, Sheng.,Liu, Liu.,Xu, Zhengxuan.,Li, Shaojun.,...&Hou, Jin.(2025).Knowledge-Based Machine learning for Real-Time rock strength testing while Drilling: Bridging Simulation and Reality.MEASUREMENT,246,17. |
| MLA | Bai, Jun,et al."Knowledge-Based Machine learning for Real-Time rock strength testing while Drilling: Bridging Simulation and Reality".MEASUREMENT 246(2025):17. |
入库方式： OAI收割
来源：武汉岩土力学研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。