基于ICEEMDAN与Attention-LSTM的矿山边坡位移预测

doi:10.12068/j.issn.1005-3026.2025.20240197

摘要/Abstract

摘要：

为了提高矿山边坡位移预测的精度，提出了1种基于改进自适应噪声完备集合经验模态分解（ICEEMDAN）、最小二乘拟合与融入注意力机制的长短时记忆网络的矿山边坡位移预测模型.首先，针对矿山边坡位移监测数据的时序性、非线性等特点，引入改进的模态分解方法将累积位移进行时频分解，获得趋势项、周期项和随机项，有效降低了数据复杂度.其次，针对趋势项预测，采用最小二乘拟合方法建立三次多项式回归预测模型；针对周期项预测，引入注意力机制区分不同时刻位移数据的重要程度，有效捕捉了长时间位移序列内部的依赖关系.最终，通过整合趋势项和周期项预测结果，获得矿山边坡累积位移的预测结果.以西藏巨龙铜矿边坡为例，测试了该模型的性能.结果表明，提出的边坡位移预测模型的均方根误差和平均绝对百分比误差分别为5.99 mm和5.94%，与传统长短时记忆网络模型相比，分别下降了51.30%和55.17%，预测精度显著提高.

关键词: 位移预测, 露天矿, 滑坡, 长短时记忆网络, 注意力机制

Abstract:

In order to improve the accuracy of mine slope displacement prediction， a mine slope displacement prediction model based on an improved complete ensemble empirical mode decomposition with adaptive noise （ICEEMDAN）， least squares fitting method and long short-term memory network integrated into the attention mechanism was proposed. Firstly， considering the temporal and nonlinear characteristics of the mining slope displacement monitoring data， an improved mode decomposition method was employed to perform the time-frequency decomposition of cumulative displacement， resulting in trend， periodic， and random components， thereby effectively reducing the data complexity. Secondly， to predict the trend component， a cubic polynomial regression prediction model was developed by using the least squares fitting method. To predict the periodic component， an attention mechanism was introduced to distinguish the importance of displacement data at different times， which effectively captured the internal dependencies within the long-term displacement sequences. Finally， the predicted results of the trend and periodic components were integrated to obtain the cumulative displacement of the mining slope. Taking the slope of Julong Copper Mine in Xizang as an example， the performance of the proposed method was tested. The results demonstrate that the proposed slope displacement prediction model achieves a root mean square error （RMSE） of 5.99 mm and a mean absolute percentage error （MAPE） of 5.94%. The RMSE and MAPE decrease by 51.30% and 55.17% compared with the traditional LSTM model， respectively. These results highlight the significant improvement in prediction accuracy achieved by the proposed method.

Key words: displacement prediction, open-pit mining, landslide, long short-term memory（LSTM）network, attention mechanism

中图分类号:

TD 804

李荟, 韩晓飞, 朱万成, 毛嘉石. 基于ICEEMDAN与Attention-LSTM的矿山边坡位移预测[J]. 东北大学学报（自然科学版）, 2025, 46(7): 163-170.

Hui LI, Xiao-fei HAN, Wan-cheng ZHU, Jia-shi MAO. Mine Slope Displacement Prediction Based on ICEEMDAN and Attention-LSTM[J]. Journal of Northeastern University(Natural Science), 2025, 46(7): 163-170.

图/表 9

图1 长短时记忆网络结构图

Fig.1 Structure diagram of the long short-term memory network

图2 注意力机制结构图

Fig.2 Structure diagram of the attention mechanism

图3 Attention-LSTM结构图

Fig.3 Structure diagram of Attention-LSTM

图4 累积位移预测流程图

Fig.4 Flowchart of cumulative displacement prediction

图5 监测点累积位移及累积位移分解（a）—累积位移；（b）—趋势项位移；（c）—周期项位移；（d）—随机项位移.

Fig.5 Cumulative displacement of monitoring points and decomposition of cumulative displacement

图6 趋势项位移预测

Fig.6 Prediction of trend displacement

图7 周期项位移预测

Fig.7 Prediction of periodic displacement

图8 累积位移预测

Fig.8 Prediction of cumulative displacement

表1 边坡累积位移预测结果 (displacement)

Table 1 Prediction results of slope cumulative

指标	LSTM	ICEEMDAN-CPF-LSTM	ICEEMDAN-CPF-Attention-LSTM
$e M A P E$ /%	13.250	7.820	5.940
$e R M S E$ /mm	12.300	7.990	5.990
$R 2$	0.800	0.905	0.964

表1 边坡累积位移预测结果 (displacement)

Table 1 Prediction results of slope cumulative

指标	LSTM	ICEEMDAN-CPF-LSTM	ICEEMDAN-CPF-Attention-LSTM
$e M A P E$ /%	13.250	7.820	5.940
$e R M S E$ /mm	12.300	7.990	5.990
$R 2$	0.800	0.905	0.964

参考文献 23

[1]	杨天鸿，王赫，董鑫，等.露天矿边坡稳定性智能评价研究现状、存在问题及对策［J］.煤炭学报，2020，45（6）：2277-2295.
	Yang Tian-hong， Wang He， Dong Xin， et al. Current situation，problems and countermeasures of intelligent evaluation of slope stability in open pit［J］. Journal of China Coal Society， 2020，45（6）：2277-2295.
[2]	李荟，韩晓飞，朱万成，等.基于多源信息融合的矿山边坡滑坡灾害研究现状与展望［J］.工矿自动化，2024，50（6）：6-15.
	Li Hui， Han Xiao-fei， Zhu Wan-cheng， et al. Current status and prospects of research on landslide disasters in mine slopes based on multi-source information fusion［J］. Journal of Mine Automation， 2024，50（6）：6-15.
[3]	Liu K M， Li H， Pang S H， et al. Numerical simulation analysis of slope instability and failure of limestone mine in Weibei［J］. Advances in Civil Engineering， 2021， 2021（1）： 5991348.
[4]	Azarfar B， Ahmadvand S， Sattarvand J， et al. Stability analysis of rock structure in large slopes and open-pit mine： numerical and experimental fault modeling［J］. Rock Mechanics and Rock Engineering， 2019， 52（12）： 4889-4905.
[5]	Zhang K， Cao P， Meng J J， et al. Modeling the progressive failure of jointed rock slope using fracture mechanics and the strength reduction method［J］. Rock Mechanics and Rock Engineering， 2015， 48（2）： 771-785.
[6]	Jiang S， Liu H S， Lian M J， et al. Rock slope displacement prediction based on multi-source information fusion and SSA-DELM model［J］. Frontiers in Environmental Science， 2022， 10： 982069.
[7]	Dai F， Zhu W C， Ren M， et al. Landslide displacement prediction based on spatio-temporal association rule mining between target case and similar cases［J］. Stochastic Environmental Research and Risk Assessment， 2023， 37（11）： 4229-4247.
[8]	Hao Y K， Huang B B， Sulowicz M. A practical prediction model for surface deformation of open-pit mine slopes based on artificial intelligence［J］. Elektronika Ir Elektrotechnika， 2024， 30（3）： 46-53.
[9]	Niu H L， Xu K L， Liu C. A decomposition-ensemble model with regrouping method and attention-based gated recurrent unit network for energy price prediction［J］. Energy， 2021， 231： 120941.
[10]	Yang B B， Yin K L， Lacasse S， et al. Time series analysis and long short-term memory neural network to predict landslide displacement［J］. Landslides， 2019， 16（4）： 677-694.
[11]	Lin Z A， Sun X Y， Ji Y F. Landslide displacement prediction model using time series analysis method and modified LSTM model［J］. Electronics， 2022， 11（10）： 1519.
[12]	Liu Z Q， Guo D， Lacasse S， et al. Algorithms for intelligent prediction of landslide displacements［J］. Journal of Zhejiang University： Science A， 2020， 21（6）： 412-429.
[13]	Yang J F， Ren Y Q， Chai J， et al. Adit deformation prediction based on ICEEMDAN dispersion entropy and LSTM-BP［J］. Optical Fiber Technology， 2023， 79： 103364.
[14]	赵二峰，李章寅，袁冬阳.基于双阶段注意力机制的大坝变形深度学习预测模型［J］.河海大学学报（自然科学版），2023，51（6）：44-52.
	Zhao Er-feng， Li Zhang-yin， Yuan Dong-yang. Deep learning model for deformation prediction of dam based on dual-stage attention mechanism ［J］. Journal of Hohai University（Natural Sciences）， 2023，51（6）：44-52.
[15]	张振坤，张冬梅，李江，等. 基于多头自注意力机制的LSTM-MH-SA滑坡位移预测模型研究［J］. 岩土力学， 2022，43（sup2）： 477-486， 507.
	Zhang Zhen-kun， Zhang Dong-mei， Li Jiang， et al. Study on LSTM-MH-SA landslide displacement prediction model based on multi-head self-attention mechanism［J］. Rock and Soil Mechanics， 2022， 43（sup2）： 477-486， 507.
[16]	唐宇峰，陈星红，蔡宇，等. 基于时序分解和SSA-LSTM-Attention模型的尾矿坝位移预测［J］. 科学技术与工程， 2023，23（29）： 12753-12759.
	Tang Yu-feng， Chen Xing-hong， Cai Yu， et al. Tailings dam displacement prediction based on time-series decomposition and SSA-LSTM-Attention model［J］. Science Technology and Engineering， 2023，23（29）： 12753-12759.
[17]	曹鑫宇，朱琳，宫辉力，等.AM-LSTM网络的北京平原东部地面沉降模拟［J］.遥感学报，2022，26（7）：1302-1314.
	Cao Xin-yu， Zhu Lin， Gong Hui-li， et al. Land subsidence simulation in the east of Beijing plain based on the AM-LSTM Network［J］. National Remote Sensing Bulletin， 2022，26（7）：1302-1314.
[18]	Li L M， Wang C Y， Wen Z Z， et al. Landslide displacement prediction based on the ICEEMDAN， ApEn and the CNN-LSTM models［J］. Journal of Mountain Science， 2023， 20（5）： 1220-1231.
[19]	Sherstinsky A. Fundamentals of recurrent neural network （RNN） and long short-term memory （LSTM） network［J］. Physica D： Nonlinear Phenomena， 2020， 404： 132306.
[20]	刘建伟，刘俊文，罗雄麟.深度学习中注意力机制研究进展［J］.工程科学学报，2021，43（11）：1499-1511.
	Liu Jian-wei， Liu Jun-wen， Luo Xiong-lin. Research progress in attention mechanism in deep learning［J］. Chinese Journal of Engineering， 2021，43（11）：1499-1511.
[21]	Brauwers G， Frasincar F. A general survey on attention mechanisms in deep learning［J］. IEEE Transactions on Knowledge and Data Engineering， 2023， 35（4）： 3279-3298.
[22]	Fu X G， Wei Y H， Su Y， et al. Shear wave velocity prediction based on the long short-term memory network with attention mechanism［J］. Applied Sciences， 2024， 14（6）： 2489.
[23]	杨背背，殷坤龙，杜娟.基于时间序列与长短时记忆网络的滑坡位移动态预测模型［J］.岩石力学与工程学报，2018，37（10）：2334-2343.
	Yang Bei-bei， Yin Kun-long， Du Juan. A model for predicting landslide displacement based on time series and long and short term memory neural network［J］. Chinese Journal of Rock Mechanics and Engineering， 2018，37（10）：2334-2343.