Preliminary damage detection and rockburst early warning for coal-rock composite based on PCA-DBSCAN

Abstract

Rockbursts frequently occur during deep mining. They are catastrophic failures of the coal-rock composite (CRC); however, identifying their precursory signals remains challenging. This study presents a principal component analysis (PCA)-density-based spatial clustering of applications with noise (DBSCAN)-assisted workflow for warning-time analysis of CRCs. Conventional uniaxial and cyclic loading-unloading tests were conducted on CRC with four different roof rock lithologies. PCA was used to reduce redundancy among multiple acoustic emission (AE) parameters, and DBSCAN was used to organize AE samples in the reduced feature space for subsequent warning-time analysis. The results indicate an inflection point in the axial strain rate before failure, which may reflect a transition in the deformation state of the composite near the critical stage. AE signal evolution shows distinct stages, with AE count and AE energy rising sharply near failure. The PCA-DBSCAN-assisted workflow identifies different warning-related patterns: responses in the coal body are more dispersed and appear earlier, whereas responses in the roof rock are more concentrated closer to failure. The workflow provides a structured basis for organizing correlated AE responses and comparatively analyzing warning-related evolution in the coal body and the roof rock. Importantly, warning time is affected by roof lithology and generally appears earlier in the coal body than in the roof rock. These findings provide a laboratory-scale basis for comparative warning analysis of CRCs under different lithological conditions.

Keywords

Coal-rock composite / Acoustic emission / Principal component analysis / Density-based spatial clustering of applications with noise / Early warning methodology

References

[1] Skoczkowski T, Bielecki S, Kochan´ski M, Korczak K. Climate—change induced uncertainties, risks and opportunities for the coal—based region of Silesia: Stakeholders’ perspectives. Environ Innov Soc Transitions 2020; 35: 460-81.

[2] Andreoni V. Drivers of coal consumption changes: a decomposition analysis for Chinese regions. Energy 2022; 242: 122975.

[3] Chai YJ, Dou LM, Cai W, Małkowski P, Li XW, Gong SY, et al. Experimental investigation into damage and failure process of coal—rock composite structures with different roof lithologies under mining—induced stress loading. Int J Rock Mech Min Sci 2023; 170: 105479.

[4] Curran G. Coal, climate and change: the narrative drivers of Australia’s coal economy. Energy Res Soc Sci 2021; 74: 101955.

[5] Erdogan S, Pata UK, Alola AA, Echebiri C. On cutting coal dependency for establishing a cleaner energy mix: a proposal for energy transition strategy in South Africa. Energy Strategy Rev 2025; 58: 101694.

[6] Zhao TB, Guo WY, Zhang DX, Tan YL, Yin YC, Tan Y, et al. Theoretical framework for stress relief—support reinforcement cooperative control of rock bursts in deep coal mining. Geohazard Mech 2024; 2(1): 49-57.

[7] Mark C. Coal bursts that occur during development: a rock mechanics enigma. Int J Min Sci Technol 2018;28(1):35—42.

[8] Wang SF, Huang LQ, Li XB. Analysis of rockburst triggered by hard rock fragmentation using a conical pick under high uniaxial stress. Tunn Undergr Space Technol 2020; 96: 103195.

[9] Yue WT, Wang EY, Feng XJ, Tan TJ, Zhang L, Chen D, et al. Homogeneity—dependent fracture behavior and instability mechanism of composite coal—rock: Insights from three—point bending tests. Int J Min Sci Technol 2025; 35(6): 913-32.

[10] Li H, Ma HL, Yang CH, Zhao K, Hu ZW, Daemen JJK. Acoustic emission characteristics of rock salt under multi—stage cyclic loading. Int J Fatigue 2023; 176: 107911.

[11] Ye CF, Chen HJ, Li CB, Wu F, Xie HP. Prediction of continuous—discontinuous failure of granite subjected to direct shearing using acousto—optical—mechanical scheme. J Rock Mech Geotech Eng 2026. in press.

[12] Gao RB, Wu F, Li CB, Ye CF, He QC, Xie HP. Damage evolution and failure modes of coal—concrete composites with varying height ratios under cyclic loading. Int J Min Sci Technol 2025; 35(7): 1093—106.

[13] Chen TX, Xu SD, He BG, Li YH, Lei H. Mechanical response of tunnel surrounding rock under blasting disturbance with different stress conditions based on acoustic emission and digital image correlation. Eng Geol 2025; 357: 108354.

[14] Ye CF, Li CB, Chen HJ, Wu F, Chu P, Xie HP. Nonlinear progressive failure mechanism and shear strength model of deeply buried Jinping marble under direct shear. Int J Rock Mech Min Sci 2026; 198: 106380.

[15] Gee M, Roshanmanesh S, Hayati F, Papaelias M. Multi—variant damage assessment in composite materials using acoustic emission. Sensors 2025; 25(12): 3795.

[16] Mei JJ, Xiao JZ, Yang HF, Yang QM, Wei TC, Zhou YW. Acoustic emission monitoring of bond failure between carbon fiber—reinforced coral concrete and stainless steel reinforcement under lateral constraints. Eng Fail Anal 2025; 182: 110075.

[17] Li CB, Yang DC, Xie HP, Ren L, Wang J. Size effect of fracture characteristics for anisotropic quasi—brittle geomaterials. Int J Min Sci Technol 2023; 33(2): 201—13.

[18] Benzon HH, Mielke A, Skovborg Ritschel TK, McGugan M, Branner K, Chen X. Acoustic emission data analytics on delamination growth in a wind turbine blade under full—scale cyclic testing. Measurement 2025; 242: 115822.

[19] Pang YF, Cao Z, Zhang XQ, Xu LJ. Dual—comb spectrum SNR enhancement for accurate gas temperature measurement through signal—noise decorrelation and principal component analysis. Opt Laser Technol 2026; 194: 114498.

[20] Zhang JM, Wang JR, Long SG, Liu GY, Wang L, Yang D. Vertical partitioned high—dimensional data publishing with differential privacy principal component analysis. Neurocomputing 2025; 655: 131415.

[21] Rodrigues da Silva E, de Andrade Oliveira Marques J, Bezerra de Araújo Medeiros RL, de Araújo Melo DM, Francisco Alves JL, Braga RM. Optimizing hydrocarbon yield from co—pyrolysis of Agave sisalana bagasse and multilayer plastic film wastes for sustainable fuel production using simplex—centroid mixture design and principal component analysis. Renew Energy 2026; 256: 124579.

[22] Gao AS, Qi CZ, Shan RL, Wang CL, Kocharyan GG. Identification and early warning method of key disaster—causing factors of AE signals for red sandstone by principal component analysis method. Ain Shams Eng J 2023; 14(10): 102205.

[23] Cao KW, Xu YJ, Khan NM, Li XC, Cui RY, Hussain S, et al. A comprehensive model for evaluating infrared radiation and acoustic emission characteristics of sandstone fracture. Eng Fract Mech 2023; 283: 109217.

[24] Wang Y, Jie G, Na W, Chao Y, Li Z, Jun CL. Identification of the concrete damage degree based on the principal component analysis of acoustic emission signals and neural networks. Mater Test 2020; 62(5): 517—24.

[25] Ester M, Kriegel HP, Sander J, Xu X. A density—based algorithm for discovering clusters in large spatial databases with noise. In: Proceedings of the Second International Conference on Knowledge Discovery and Data Mining. Portland: AAAI Press. 1996.p. 226-31.

[26] Rui YC, Zhou ZL, Cai X, Dong LJ. A novel robust method for acoustic emission source location using DBSCAN principle. Measurement 2022; 191: 110812.

[27] de la Selle T, Weiss J, Deschanel S. Acoustic multiplets detection based on DBSCAN and cross—correlation. Mech Syst Signal Process 2024; 211: 111149.

[28] Wang CL, Zhou BK, Zhu CY, Li CF, Sun L. Crack nucleation mechanism of rock fracture. Eng Fract Mech 2025; 325: 111328.

[29] Lu J, Huang G, Gao H, Li X, Zhang DM, Yin GZ. Mechanical properties of layered composite coal—rock subjected to true triaxial stress. Rock Mech Rock Eng 2020; 53(9): 4117—38.

[30] Li DX, Wang EY, Kong XG, Jia HS, Wang DM, Muhammad A. Damage precursor of construction rocks under uniaxial cyclic loading tests analyzed by acoustic emission. Constr Build Mater 2019; 206: 169-78.

[31] Shah KR, Labuz JF. Damage mechanisms in stressed rock from acoustic emission. J Geophys Res Solid Earth 1995; 100(B8): 15527—39.

[32] Ye CF, Li CB, Chen HJ, Wu F, Chu P, Xie HP. Progressive failure mechanism and shear strength model of granite under cyclic direct shear. Rock Mech Rock Eng 2026; 59(2): 1455-82.

[33] Meng FZ, Zhou H, Wang ZQ, Zhang LM, Kong L, Li SJ, et al. Experimental study on the prediction of rockburst hazards induced by dynamic structural plane shearing in deeply buried hard rock tunnels. Int J Rock Mech Min Sci 2016; 86: 210-23.

[34] Ran QC, Liang YP, Ye CF, Song X, Ma TF, Chen ZH. Mechanical deterioration and crack propagation characteristics of mudstone under cyclic loading: insights from acoustic emission fractal analysis. Rock Mech Rock Eng 2025: 1-28.

[35] Xie PY, Chen WZ, Zhao WS, Gao H. Acquisition of acoustic emission precursor information for rock masses with a single joint based on clustering—convolutional neural network method. J Rock Mech Geotech Eng 2024; 16(12): 5061-76.

[36] Xue CW, Liu T, Deng LB, Gu W, Zhang BW. Improved differential evolution algorithm based on PCA—DBSCAN classification. Mod Electron Tech 2024; 47(16): 171-9. in Chinese.

[37] Jia ZC, Wang Y, Wang JH, Pei QY, Zhang YQ. Rockburst intensity grade prediction based on data preprocessing techniques and multi—model ensemble learning algorithms. Rock Mech Rock Eng 2024; 57(7): 5207—27.

[38] Ma K, Xie HB, Ren FQ, Chang Y. Rockburst early—warning method based on time series prediction of multiple acoustic emission parameters. Tunn Undergr Space Technol 2024; 153: 106060.

[39] Wang JX, Wang EY, Yang WX, Li BL, Li ZH, Liu XF. Rock burst monitoring and early warning under uncertainty based on multi—information fusion approach. Measurement 2022; 205: 112188.

[40] Liu CY, Chen QF. Deep learning—based multi—parameter early warning model under true triaxial conditions. Eng Geol 2023; 319: 107111.

[41] Ran QC, Liang YP, Zou QL, Ye CF, Chen ZH, Ma TF, et al. Mechanical behavior and acoustic emission characteristics of initially damaged coal under triaxial cyclic loading and unloading. J Rock Mech Geotech Eng 2026; 18(1): 230-45.

[42] Xin TY, Ji Y, Wang JG, Bai G, Sun VJ, Li JY, et al. Study on deformation characteristics and damage evolution of water—bearing coal under cyclic loading. Results Eng 2025; 27: 105906.

[43] Shan TC, Li ZH, Wang EY, Jia HS, Zhang X, Zhang QM, Wang XR, Niu Y, Deng SS. Identification of coal—rock interface under uniaxial compression using electric potential method. J Rock Mech Geotech Eng 2026. in press.

[44] Carpinteri A, Lacidogna G, Accornero F, Mpalaskas AC, Matikas TE, Aggelis DG. Influence of damage in the acoustic emission parameters. Cem Concr Compos 2013; 44: 9-16.

[45] Aggelis DG, Mpalaskas AC, Ntalakas D, Matikas TE. Effect of wave distortion on acoustic emission characterization of cementitious materials. Constr Build Mater 2012; 35: 183-90.

[46] Ren FQ, Gao ZY, Ma K, Yang S. Rockburst probability early warning method based on integrated infrared temperature and acoustic emission parameters. Int J Rock Mech Min Sci 2025; 189: 106097.