(1. 中国矿业大学（北京） 力学与土木工程学院，北京 100083；2. Research Institute, ESTP-Grande École d’Ingénieurs de la Construction, Cachan 94234, France；3. 中国矿业大学(北京) 采矿岩石力学国际创新中心，北京 100083)
Prediction of elastic modulus in heterogeneous rocks through fusion of multi-source images and artificial intelligence
(1. School of Mechanics and Civil Engineering, China University of Mining and Technology(Beijing), Beijing 100083, China；2. Research Institute, ESTP-Grande École d’Ingénieurs de la Construction, Cachan 94234, France；3. International Innovation Center for Mining Rock Mechanic, China University of Mining and Technology(Beijing), Beijing 100083, China)
Accurate measurement of the elastic modulus of rock is a key component in geotechnical engineering construction design and stability evaluation. However, in deep high-stress environments and brittle formations with well-developed weak structures, obtaining intact and standardized rock samples for direct measurement of elastic modulus is often challenging. This study proposes a method for predicting elastic modulus by leveraging artificial intelligence to build a bridge between microstructural characteristics and elastic modulus, requiring only microstructure images as input to efficiently infer the modulus. The implementation process consists of two stages: first, an improved DeepLabv3+ model is employed to perform semantic segmentation experiments on a multi-source image dataset, generating microstructural characterization images; second, a RockNet-E model is constructed to establish a mapping relationship between the characterization images and the elastic modulus, enabling accurate prediction. Research results show that the improved DeepLabv3+ model achieves mean intersection over union (IoU) of 87.7%, 97.5% and 87.9% for quartz, feldspar, and mica, respectively, outperforming mainstream models such as DeepLabv3+, PSPNet, and UNet. On the validation set, the RockNet-E model achieves mean absolute error (MAE), root mean square error (RMSE), mean absolute percentage error (MAPE) and R-squared (R2) values of 1.20 GPa, 1.51 GPa, 1.83% and 0.61, respectively. The elastic modulus prediction errors are all below 5%, and the predicted value distribution matches the experimental results. This study provides a low-cost and highly efficient approach for intelligent, image-driven prediction of rock mechanical parameters.
[1] 谢和平，张 茹，张泽天，等. 深地科学与深地工程技术探索与思考[J]. 煤炭学报，2023，48(11)：3 959–3 978.(XIE Heping，ZHANG Ru，ZHANG Zetian，et al. Reflections and explorations on deep earth science and deep earth engineering technology[J]. Journal of China Coal Society，2023，48(11)：3 959–3 978.(in Chinese))
[2] 谢和平. “深部岩体力学与开采理论”研究构想与预期成果展望[J]. 工程科学与技术，2017，49(2)：1–16.(XIE Heping. Framework and anticipated results of deep rock mechanics and mining theory[J]. Advanced Engineering Science，2017，49(2)：1–16. (in Chinese))
[3] 田文岭，杨圣奇，黄彦华，等. 花岗岩高温高压损伤破裂细观机制模拟研究[J]. 岩石力学与工程学报，2022，41(9)：1 810–1 819. (TIAN Wenling，YANG Shengqi，HUANG Yanhua，et al. Meso-fracture mechanism of granite specimens under high temperature and confining pressure by numerical simulation[J]. Chinese Journal of Rock Mechanics and Engineering，2022，41(9)：1 810–1 819.(in Chinese))
[4] LIU Z D，LI D Y，LIU Y P，et al. Prediction of uniaxial compressive strength of rock based on lithology using stacking models[J]. Rock Mechanics Bulletin，2023，2(4)：100081.
[5] MOMENI E，ARMAGHANI D J，HAJIHASSANI M，et al. Prediction of uniaxial compressive strength of rock samples using hybrid particle swarm optimization-based artificial neural networks[J]. Measurement，60：50–63.
[6] GUERY A C，CORMERY F，SHAO J F，et al. Application of a micromechanical model to cavity excavation analysis in argillite[J]. International Journal of Rock Mechanics and Mining Sciences，2009，46(5)：905–917.
[7] 时 贤，蒋 恕，卢双舫，等. 利用纳米压痕实验研究层理性页岩岩石力学性质—以渝东南酉阳地区下志留统龙马溪组为例[J]. 石油勘探与开发，2019，46(1)：155–164.(SHI Xian，JIANG Shu，LU Shuangfang，et al. Investigation of mechanical properties of bedded shale by nanoindentation tests：A case-study on Lower Silurian Longmaxi Formation of Youyang area in southeast Chongqing，China[J]. Petroleum Exploration and Development，2019，46(1)：155–164.(in Chinese))
[8] 张 帆，郭翰群，赵建建，等. 花岗岩微观力学性质试验研究[J]. 岩石力学与工程学报，2017，36(增2)：3 864–3 872.(ZHANG Fan，GUO Hanqun，ZHAO Jianjian，et al. Experimental study of micro–mechanical properties of granite[J]. Chinese Journal of Rock Mechanics and Engineering，2017，36(Supp.2)：3 864–3 872.(in Chinese))
[9] 徐鼎平，柳秀洋，徐怀胜，等. 深埋花岗岩细观力学特性纳米压痕试验及参数均质化研究[J]. 中南大学学报：自然科学版，2021，52(8)：2 761–2 771.(XU Dingping，LIU Xiuyang，XU Huaisheng，et al. Meso-mechanical properties of deep granite using nanoindentation test and homogenization approach[J]. Journal of Central South University：Science and Technology，2021，52(8)：2 761–2 771.(in Chinese))
[10] YUE Z Q，CHEN S，THAM L G. Finite element modeling of geomaterials using digital image processing[J]. Computers and Geotechnics，30(5)：375–397.
[11] 段永婷，冯夏庭，李 晓. 页岩细观矿物条带对其宏观破坏模式的影响研究[J]. 岩石力学与工程学报，2021，40(1)：43–52.(DUAN Yongting，FENG Xiating，LI Xiao. Study on the influence of meso-mineral bands on macroscopic failure modes of shale[J]. Chinese Journal of Rock Mechanics and Engineering，2021，40(1)：43–52.(in Chinese))
[12] QIN X R，ZHANG H M，XU R P，et al. Experimental investigation on the failure mechanism of grouted rock mass: Mesostructure and macroscopic mechanical behavior[J]. Engineering Failure Analysis，2024，161：108304.
[13] ZUO J P，WANG J T，SUN Y J，et al. Effects of thermal treatment on fracture characteristics of granite from Beishan, a possible high-level radioactive waste disposal site in China[J]. Engineering Fracture Mechanics，2017，182：425–437.
[14] LI C X，LIU Y S，LI L S，et al. Texture-based segmentation of SEM images of shale rocks and estimation of meso-scale elastic modulus by 2D FEM[J]. Rock Mechanics and Rock Engineering，2025，58：6 475–6 491.
[15] 李明耀，彭 磊，左建平，等. 基于 DIP-FFT 数值方法的花岗岩多尺度力学特性研究[J]. 岩石力学与工程学报，2022，41(11)：2 254–2 267.(LI Mingyao，PENG Lei，ZUO Jianping，et al. Study on multi-scale mechanical properties of granite based on DIP-FFT numerical method[J]. Chinese Journal of Rock Mechanics and Engineering，2022，41(11)：2 254–2 267.(in Chinese))
[16] WANG F，KONIETZKY H，HERBST M，et al. Mechanical responses of grain-based models considering different crystallographic spatial distributions to simulate heterogeneous rocks under loading[J]. International Journal of Rock Mechanics and Mining Sciences，2022，151：105036.
[17] 周 中，张俊杰，龚琛杰，等. 基于深度语义分割的隧道渗漏水智能识别[J]. 岩石力学与工程学报，2022，41(10)：2 082–2 093.(ZHOU Zhong，ZHANG Junjie，GONG Chenjie，et al. Automatic identification of tunnel leakage based on deep semantic segmentation[J]. Chinese Journal of Rock Mechanics and Engineering，2022，41(10)：2 082–2 093.(in Chinese))
[18] ZHANG P，TANG K K，CHEN G G，et al. Multimodal data fusion enhanced deep learning prediction of crack path segmentation in CFRP composites[J]. Composites Science and Technology，2024，257：110812.
[19] LI X，LIU Z L，CUI S Q，et al. Predicting the effective mechanical property of heterogeneous materials by image based modeling and deep learning[J]. Computer Methods in Applied Mechanics and Engineering，2019，347：735–53.
[20] WU Z J，WU Y，WENG L，et al. Machine learning approach to predicting the macro-mechanical properties of rock from the meso-mechanical parameters[J]. Computers and Geotechnics，2024，166：105933.
[21] 张 帆，胡 维，郭翰群，等. 热处理后花岗岩纳米压痕试验研究[J]. 岩土力学，2018，39(增1)：235–243.(ZHANG Fan，HU Wei，GUO Hanqun，et al. Nanoindentation tests on granite after heat treatment[J]. Rock and Soil Mechanics，2018，39(Supp.1)：235–243.(in Chinese))
[22] 杨振琦，王述红，孟嫣然，等. 基于 CT 扫描的花岗岩三维数值试件重构模型及应用[J]. 固体力学学报，2017，38(6)：591–600. (YANG Zhenqi，WANG Shuhong，MENG Yanran，et al. The 3D numerical specimen of granite based on CT technology and application[J]. Chinese Journal of Solid Mechanics，2017，38(6)：591–600.(in Chinese))
[23] 毛伟泽，吕 庆，郑 俊，等. 基于CT图像的花岗岩矿物组分与细观结构分析[J]. 工程地质学报，2022，30(1)：216–222.(MAO Weize，LV Qing，ZHENG Jun，et al. Analysis of mineral composition and meso-structure of granite using CT images[J]. Journal of Engineering Geology，2022，30(1)：216–222.(in Chinese))
[24] 李 博，梁秦源，周 宇，等. 基于CT-GBM重构法的花岗岩裂纹扩展规律研究[J]. 岩石力学与工程学报，2022，41(6)：1 114–1 125 (LI Bo，LIANG Qinyuan，ZHOU Yu，et al. Research on crack propagation law of granite based on CT-GBM reconstruction method [J]. Chinese Journal of Solid Mechanics，2022，41(6)：1 114–1 125.(in Chinese))
[25] PENG L，LI M Y，ZUO J P，et al. Determination of the REV size for heterogeneous rocks with different grain sizes: Deep learning and numerical approaches[J]. International Journal of Rock Mechanics and Mining Sciences，2024，183：105940.
[26] LI M Y，PENG L，LIU D J，et al. Microstructure effect of mechanical and cracking behaviors on brittle rocks using image-based fast Fourier transform method[J]. Journal of Rock Mechanics and Geotechnical Engineering，2025，17：399–413.
2026, Vol. 45 
