| 185 | 1 | 61 |
| 下载次数 | 被引频次 | 阅读次数 |
尾矿坝沉降变形具有复杂时空耦合特性,传统方法难以有效捕捉监测点间空间关联与时间动态演化。针对红岩沟尾矿坝的20个监测点(初期坝:J1/G1/X1/G2/J2;堆积坝:一级子坝D1/D2/D3/D4/D5,二级子坝C1/C2/C3/C4;副坝:G9/G10/G11/J9/J10/J11),提出基于图卷积神经网络–长短期记忆网络(GCN–LSTM)的时空混合预测模型。首先采用小波阈值法预处理沉降数据,基于皮尔逊相关系数构建监测点空间关联的加权无向图;进而利用GCN模块提取空间拓扑特征,耦合LSTM模块学习时间依赖关系,建立时空联合预测框架;最终通过Adam优化器调优超参数。试验结果表明:模型预测精度显著优于传统方法 (RMSE=0.023 35,R=0.995 11),在结构异质性区域(初期坝/堆积坝/副坝)均能精准捕捉沉降趋势;消融试验中,其R较单一LSTM、GNN模型分别提高0.189 14与0.347 27,该模型通过融合空间关联与时间动态特性,实现了尾矿坝多点沉降高精度预测,为安全状态评估与溃坝风险预警提供了可靠技术支撑。
Abstract：The settlement deformation of tailing dams exhibits complex spatiotemporal coupling characteristics,making it difficult for using traditional methods to effectively capture spatial correlations between monitoring points and the temporal dynamic evolution. For 20 monitoring points of the Hongyangou tailings dam(Initial dam: J1/G1/X1/G2/J2; Accumulation dam: first-stage sub-dam D1/D2/D3/D4/D5, second-stage sub-dam C1/C2/C3/C4; Auxiliary dam: G9/G10/G11/J9/J10/J11), a spatiotemporal hybrid prediction model based on a Graph Convolutional Network-Long Short-Term Memory(GCN-LSTM) architecture is proposed. First, the settlement data are preprocessed using the wavelet threshold denoising method, and a weighted undirected graph of spatial correlations among monitoring points is constructed based on the Pearson correlation coefficient.Subsequently, the GCN module is employed to explore the spatial topological features, which are then coupled with the LSTM module to learn temporal dependencies, thereby establishing a joint spatiotemporal prediction framework. Finally, the Adam optimizer is used to fine-tune the hyperparameters. Experimental results demonstrate that the proposed model achieves significantly higher prediction accuracy than those of traditional methods(RMSE = 0.023 35, R = 0.995 11) and can accurately capture settlement trends even in structurally heterogeneous regions(initial dam/accumulation dam/auxiliary dam). In ablation experiments, the R value of the proposed model increased by 0.189 14 and 0.347 27 when compared with the standalone LSTM and GNN models, respectively. By integrating spatial correlation and temporal dynamic characteristics, the proposed model enables high-precision multi-point settlement prediction of tailing dams, providing a reliable technical basis for safety assessment and risk early warning dam failure.
[1]DU Z, GE L, NG H A, et al. Risk assessment for tailings dams in Brumadinho of Brazil using InSAR time series approach[J]. Science of the Total Environment, 2020, 717:137125-137138.
[2]李辉,易富,张佳.基于因素空间的尾矿坝稳定性综合评价[J].中国安全科学学报, 2019, 29(12):28-34.LI Hui, YI Fu, ZHANG Jia. Comprehensive stability evaluation of tailing dam based on factor space[J]. China Safety Science Journal, 2019, 29(12):28-34.
[3]LI J, CHEN H, ZHOU T, et al. Tailings pond risk prediction using long short-term memory net-works[J]. IEEE Access, 2019, 7:182527-182537.
[4]沈楼燕,李连通,张超,等.我国尾矿库安全监测技术发展综述[J].有色金属工程, 2023, 13(1):121-126.SHEN Louyan, LI Liantong, ZHANG Chao, et al. Review on the development of safety monitoring technology for tailings pond in China[J]. Nonferrous Metals, 2023, 13(1):121-126.
[5]朱美宣.基于ABAQUS的高填方边坡沉降数值模拟分析[J].智能计算机与应用, 2025, 15(3):106-113.ZHU Meixuan. Numerical simulation analysis of high fill slope settlement based on ABAQUS[J]. Intelligent Computers and Applications, 2025, 15(3):106-113.
[6]张予东,马春艳.基于InSAR技术和SA–SVR算法的矿区沉降预测模型[J].金属矿山, 2020, 49(11):197-202.ZHANG Yudong, MA Chunyan. Prediction model of mining area subsidence based on InSAR technique and SASVR Algorithm[J]. Metal Mines, 2020, 49(11):197-202.
[7]黄定川,谢世成.一种基于BP神经网络的尾矿坝沉降预报方法[J].测绘工程, 2016, 25(8):53-56, 64.HUANG Dingchuan, XIE Shicheng. A way to predict the settlement of tailings dam based on BP neural network[J].Engineering of Surveying and Mapping, 2016, 25(8):53-56, 64.
[8]李如仁,孙加瑶.融合SBAS–InSAR与GS–LSTM的尾矿库沉降监测与预测[J].金属矿山, 2023, 52(1):102-109.LI Ruren, SUN Jiayao. Subsidence monitoring and prediction of tailings pond combined with SBAS-InSAR and GSLSTM[J]. Metal Mines, 2023, 52(1):102-109.
[9]刘传立,刘小生,李妍妍.基于GEP的金属矿尾矿坝变形预测模型研究[J].有色金属科学与工程, 2013, 4(6):63-68.LIU Chuanli, LIU Xiaosheng, LI Yanyan. Deformation prediction model of metal mine tailings dam based on GEP[J]. Nonferrous Metals Science and Engineering, 2013,4(6):63-68.
[10]陈娅男,李素敏,郭瑞,等.基于时序InSAR的覆砂石尾矿坝形变演化研究[J].中国安全生产科学技术, 2020,16(4):31-37.CHEN Yanan, LI Sumin, GUO Rui, et al. Study on deformation evolution of sandstone-covered tailings dam based on time series InSAR[J]. Journal of Safety Science and Technology, 2020, 16(4):31-37.
[11]杨超,杨鹏,吕文生,等.基于无人机摄影测量的尾矿坝边坡表面变形监测[J].中国安全生产科学技术, 2021,17(5):5-11.YANG Chao, YANG Peng, LYU Wensheng, et al. Surface deformation monitoring of tailings dam slope based on UAV photogrammetry[J]. Journal of Safety Science and Technology, 2021, 17(5):5-11.
[12]MA Z, MEI G, PREZIOSO E, et al. A deep learning approach using graph convolutional networks for slope deformation prediction based on time-series displacement data[J]. Neural Computing&Applications, 2021, 33(21):14441-14457.
[13]YANG Z, YI J G, ZHEN H W, et al. A tailings dam longterm deformation pre-diction method based on empirical mode decomposition and LSTM model combined with attention mechanism[J]. Water, 2022, 14(8):1229.
[14]HUANG F, HUANG J, JIANG S, et al. Landslide displacement prediction based on multivariate chaotic model and extreme learning machine[J]. Engineering Geology,2017, 218:173-186.
[15]WANG Y, LIU M, HUANG Y, et al. Knowledge-based and data-driven underground pressure forecasting based on graph structure learning[J]. International Journal of Machine Learning and Cybernetics, 2022, 15(1):3-18.
[16]LIRA H, MARTI L, SANCHEZ–Pi N. A Graph Neural Network with spatio-temporal attention for multi-sources time series data:an applocation to frost forecast[J]. Sensors, 2022, 22(4):1486-1505.
[17]MA F, GAO F, SUN J, et al. Attention Graph Convolution Network for Image Segmentation in Big SAR Imagery Data[J]. Remote Sensing, 2019, 11(21):2586-2586.
[18]JIA Nan, TIAN Xiaolin, GAO Wenxing, et al. Deep graphconvolutional generative adversarial network for semi-supervised learning on graphs[J]. Remote Sensing, 2023,15(12):1821-1832.
[19]戴健非,杨鹏,诸利一,等.集成PCA和LSTM神经网络的浸润线预测方法[J].中国安全科学学报, 2020, 30(3):94-101.DAI Jianfei, YANG Peng, ZHU Liyi, et al. A PCA–LSTM neural network-integrated method for phreatic line prediction[J]. China Safety Science Journal, 2020,30(3):94-101.
[20]YANG Ding, YE Xiaowei, ZHI Ding, et al. Short-term tunnel-settlement prediction based on Bayesian wavelet:a probability analysis method[J]. Journal of Zhejiang University-Science A, 2023, 24(11):960-977.
[21]HOCHREITER S, SCHMIDHUBER J. Long short-term memory[J]. Neural computation, 1997, 9(8):1735-1780.
[22]SHERSTINGSJY A. Fundamentals of recurrent neural network(RNN)and long short-term memory(LSTM)network[J]. Physica D:Non-linear Phenomena, 2020, 404:132306.
[23]蔡晓光,郑学鑫,黄鑫.某铜矿尾矿砂力学特性和尾矿坝变形稳定性分析[J].科学技术与工程, 2017, 17(8):128-134.CAI Xiaoguang, ZHENG Xuexin, HUANG Xin. Mechanical characteristics of copper mine tailings and analysis on deformation and stability of tailing dam[J]. Science and Technology and Engineering, 2017, 17(8):128-134.
[24]杨春和,张超,李全明,等.大型高尾矿坝灾变机制与防控方法[J].岩土力学, 2021, 42(1):1-17.YANG Chunhe, ZHANG Chao, LI Quanming, et al. Disaster mechanism and prevention methods of large-scale high tailings dam[J]. Rock and Soil Mechanics, 2021,42(1):1-17.
[25]尹光志,敬小非,魏作安,等.尾矿坝溃坝相似模拟试验研究[J].岩石力学与工程学报, 2010, 29(S2):3830-3838.YIN Guangzhi, JING Xiaofei, WEI Zuoan, et al. Experimental study of similarity simulation of Tailings DamBreaches[J]. Chinese Journal of Rock Mechanics and Engineering, 2010, 29(S2):3830-3838.
[26]杜君武,黄庆享,韦业豪,等.厚黄土层极近距离采空区下相邻工作面相向开采时序研究[J].采矿与岩层控制工程学报, 2025, 7(2):023543.DU Junwu, HUANG Qingxiang, WEI Yehao, et al. Study on time sequence for opposite mining of adjacent working faces under ultra-close coal seam goafs with thick loess layer[J]. Journal of Mining and Strata Control Engineering, 2025, 7(2):023543.
[27]龙泊含,雍睿,钟振,等.基于改进Morgenstern-Price法的露天矿山边坡稳定性分析[J].采矿与岩层控制工程学报, 2023, 5(5):053039.LONG Bohan, YONG Rui, ZHONG Zhen, et al. Slope stability analysis of open pit mine based on the improved Morgenstern-Price method[J]. Journal of Mining and Strata Control Engineering, 2023, 5(5):053039.
[28]王婕婷,张泽珑,李飞江,等.基于图神经网络的时序信号异常检测方法[J].西北大学学报(自然科学版), 2025,55(2):343-354.WANG Jieting, ZHANG Zelong, LI Feijiang, et al. A time series signal anomaly detection method based on graph neural network[J]. Journal of Northwest University(Natural Science Edition), 2025, 55(2):343-354.
[29]张懿,张向阳,卜庆为,等.厚煤切顶巷道顶板围岩支护承载稳定性分析[J].采矿与岩层控制工程学报, 2022,4(6):063014.ZHANG Yi, ZHANG Xiangyang, BU Qingwei, et al.Analysis on bearing stability of roof surrounding rock support in thick coal roof cutting roadway[J]. Journal of Mining and Strata Control Engineering, 2022, 4(6):063014.
[30]刘迪,李泽宇,卢才武,等.毛细驱动诱发的尾矿强度劣化演化规律及机制[J].安全与环境学报, 2024, 24(2):503-510.LIU Di, LI Zeyu, LU Caiwu, et al. Evolutionary law and mechanism of tailings strength deterioration under the influence of capillary[J]. Journal of Safety and Environment, 2024, 24(2):503-510.
[31]陈善雄,许锡昌,徐海滨.降雨型堆积层滑坡特征及稳定性分析[J].岩土力学, 2005, 26(S2):6-10.CHEN Shanxiong, XU Xichang, XU Haibin. Features and stability analysis of rainfall-induced colluvial landslides[J]. Rock and Soil Mechanics, 2005, 26(S2):6-10.
[32]柯丽华,张莹,李全明,等.基于EAHP的尾矿库溃坝风险多级模糊综合评价研究[J].金属矿山, 2020(11):37-43.KE Lihua, ZHANG Ying, LI Quanming, et al. Multi-level fuzzy comprehensive evaluation model based on EAHP for dam break riskof tailings pond[J]. Metal Mine,2020(11):37-43.
[33]刘迪,杨辉,卢才武,等.基于MISSA–CNN–BiLSTM模型的尾矿坝位移预测[J].中国安全科学学报, 2024,34(9):145-154.LIU Di, YANG Hui, LU Caiwu, et al. Prediction of displacement of tailings dams based on MISSA-CNN-BiLSTM model[J]. China Safety Science Journal, 2024,34(9):145-154.
基本信息:
DOI：10.13532/j.jmsce.cn10-1638/td.2025-1252
中图分类号:TD926.4
引用信息:
[1]刘迪,刘曜华,卢才武,等.基于GCN–LSTM的尾矿坝多点位沉降变形预测方法[J].采矿与岩层控制工程学报,2025,7(06):236-250.DOI:10.13532/j.jmsce.cn10-1638/td.2025-1252.
基金信息:
国家自然科学基金资助项目(52404140); 陕西省自然科学基金资助项目(S2023-JC-QN-0687); 陕西省社会科学基金资助项目(2023R035)