李果

• 1:国能神东煤炭集团有限责任公司

摘要(Abstract)：

矿井涌（突）水易造成严重伤亡与财产损失，精准预测工作面涌水量对灾害防控至关重要。为提升预测准确度并降低噪声干扰，本文提出基于变分模态分解（VMD）的混合涌水量预测模型。首先，通过VMD将涌水量数据分解为8个本征模态函数（IMF）分量以削弱噪声；随后筛选核心分量，针对不同特征选取适配模型训练，对于绝对主导分量IMF1采用Prophet模型，周期性分量IMF2、3采用长短期记忆网络（LSTM），复杂模式分量IMF4采用傅里叶特征提取结合LSTM(F-LSTM)，低频分量IMF7采用ARIMA模型。最后，按贡献率融合各核心分量预测结果得到最终值。对比实验表明，该模型平均绝对误差（MAE）、均方根误差（RMSE）、拟合系数（R~2）分别为7.52 m~3/h、10.3 m~3/h、97%，显著优于单一的Prophet、LSTM、F-LSTM及ARIMA模型，可有效抗干扰、提精度、加速率，适用于工作面涌水量预测。

关键词(KeyWords)： 涌水量预测;混合模型;变分模态分解;分量预测

Abstract:

Keywords:

基金项目(Foundation): 国能神东煤炭科技项目（E210100579）

作者(Author): 李果

参考文献(References)：

• [17]毕贵红，黄泽，赵四洪，等.基于混合分解和PCG-Bi LSTM的风速短期预测[J].太阳能学报，2024,45(1):159-170.BI G,HUANG Z,ZHAO S,et al.Short-term prediction of wind speed based on hybrid decomposition and PCG-BILSTM[J].Acta Energiae Solaris Sinica,2024,45(1):159-170.
• [18]郇嘉嘉，李代猛，杜云飞，等.基于Prophet算法和Blending集成学习的实时负荷中期预测[J].电力自动化设备，2024,44(4):178-183.HUAN J,LI D,DU Y,et al.Mid-term forecasting of real-time load based on Prophet algorithm and Blending integrated learning[J].Electric Power Automation Equipment,2024,44(4):178-183.
• [19]连会青，李启兴，王瑞，等.基于深度学习的LSTM-GRU复合模型矿井涌水量预测方法研究[J].煤矿安全，2024,55(9):166-172.LIAN H,LI Q,WANG R,et al.Research on mine water inflow prediction method of LSTM-GRU composite model based on deep learning[J].Safety in Coal Mines,2024,55(9):166-172.
• [1]国家矿山安全监察局.2023年全国矿山安全生产形势分析报告[R].北京：国家矿山安全监察局，2024.
• [2]侯恩科，席慧琴，文强，等.基于GMS的隐伏火烧区下煤层开采工作面涌水量预测[J].安全与环境学报，2022,22(5):2482-2492.HOU E,XI H,WEN Q,et al.Prediction of water inflow volume in the coal mining workforce below the concealed fire area based on GMS[J].Safety and Environmental Engineering,2022,22(5):2482-2492.
• [3]施龙青，史雅迪，李越，等.传递岩梁初次断裂工作面涌水量预测模型研究及应用[J].煤炭工程，2021,53(4):110-114.SHI L,SHI Y,LI Y,et al.Prediction model of working face water inflow when transfer rock beam initially fractures[J].Coal Engineering,2021,53(4):110-114.
• [4]刘常昊，郑万波，吴燕清，等.玉磨铁路景寨隧道富水段涌水量预测方法[J].安全与环境工程，2022,29(4):66-73.LIU C,ZHENG W,WU Y,et al.Predictio method of water inflow at water-rich section in Jingzhai tunnel of Yuxi-Mohan railway[J].Safety and Environmental Engineering,2022,29(4):66-73.
• [5]杨建，王皓，梁向阳，等.鄂尔多斯盆地北部深埋煤层工作面涌水量预测方法[J].煤田地质与勘探，2021,49(4):185-191.YANG J,WANG H,LIANG X,et al.Water inflow forecasting method of deep buried coal working face in northern Ordos Basin[J].Coal Geology&Exploration,2021,49(4):185-191.
• [6]高瑞忠，姬腾达，于婵，等.基于钻探-物探数据融合的矿井涌水量预测及效果评估[J].煤矿安全，2021,52(8):66-74.GAO R,Jl T,YU C,et al.Mine water inflow prediction and effect evaluation based on fused drilling-geophysical data[J].Safety in Coal Mines,2021,52(8):66-74.
• [7]吴卫忠，邓忠，陈余道，等.广西盘龙铅锌矿涌水量时间序列变化特征分析及ARIMA预测[J].桂林理工大学学报，2023,43(3):406-413.WU W,DENG Z,CHEN Y,et al.Characteristic analysis and ARIMAprediction of water inflow time series for Panlong lead-zinc deposit in Guangxi[J].Journal of Guilin University of Technology,2023,43(3):406-413.
• [8]施龙青，王雅茹，邱梅，等.时间序列模型在工作面涌水量预测中的应用[J].煤田地质与勘探，2020,48(3):108-115,121.SHI L,WANG Y,QIU M,et al.Application of time series model in prediction of water inflow in working face[J]Coal Geology&Exploration,2020,48(3):108-115,121.
• [9]王春刚，方刚，刘洋.榆横北区巴拉素井田不同时期矿井涌水量预测[J].煤矿安全，2020,51(6):212-217,222.WANG C,FANG G,Ll U Y.Prediction of mine water inflow at different periods in Balasu coal field of northern Yuheng mining area[J].Safety in Coal Mines,2020,51(6):212-217,222.
• [10]DRAGOMIRETSKIY K,ZOSSO D.Variational mode decomposition[J].IEEE Transactions on Signal Processing,2014,62(3):531-544.
• [11]罗灿坤，刘昊，黄鑫，等.基于VMD-LSTM-ARMA模型的径流预测[J].人民珠江，2023,44(4):96-102.LUO C,LIU H,HUANG X,et al.Runoff prediction based on VMD-LSTM-ARMA model[J].Pearl River,2023,44(4):96-102.
• [12]佟磊，张媛，吉宇豪，等.基于EMD-LSTM预测模型与夏普比率风险回溯的交易策略研究[J].中国市场，2023 (2):37-41.TONG L,ZHANG Y,JI Y,et al.Research on trading strategies based on EMD-LSTM prediction model and sharpe ratio risk backtracking[J].China Market,2023 (2):37-41.
• [13]侯恩科，夏冰冰，吴章涛，等.基于CEEMDAN-BO-Bi GRU的矿井涌水量预测研究[J].科学技术与工程，2023,23(28):12012-12019.HOU E,XIA B,WU Z,et al.Mine water inflow prediction based on CEEMDAN-BO-Bi GRU[J].Science Technology and Engineering,2023,23(28):12012-12019.
• [14]刘慧，刘桂芹，宁殿艳，等.基于VD-DBN的矿井涌水量预测方法[J]，煤田地质与勘探，2023,51(6):13-21.LU H,Ll U G,NING D,et al.Mine water inrush prediction method based on VMD-DBN model[J].Coal Geology&Exploration,2023,51(6):13-21.
• [15]丁莹莹，尹尚先，连会青，等.基于SSA-CG-Attention模型的多因素采煤工作面涌水量预测[J].煤田地质与勘探，2024,52(4):111-119.DING Y,YIN S,LIAN H,et al.Prediction of mine water inflow along mining faces using the SSA-CG-Attention multifactor model[J].Coal Geology&Exploration,2024,52(4):111-119.
• [16]李占利，邢金莎，靳红梅，等.基于CEEMD_GRU模型的矿井涌水量预测[J].北京工业大学学报，2021,47 (8):904-911.LI Z,XING J,JIN H,et al.Prediction of mine water inflow based on CEEMD GRU model[J].Journal of Beijing University of Technology,2021,47 (8):904-911.
• [20]丁莹莹，尹尚先，连会青，等.基于CEEMDAN和改进的混合时间序列模型工作面涌水量预测研究[J].中国安全生产科学技术，2024,20(3):110-117.DING Y,YIN S,LIAN H,et al.Research on water inflow prediction of working face based On CEEMDAN and improved hybrid time series model[J].Journal of Safety Science and Technology,2024,20(3):110-117.
• [21]侍建国，张亦飞.拉依达准则在处理区域水文数据异常值中的应用[J].海河水利，2016(5):49-51.SHI J,ZHANG Y.Application of Lajda’s criterion in dealing with regional hydrological data outliers[J].Haihe Water Resources,2016(5):49-51.
• [22]梁满玉，尹尚先，姚辉等.基于DRN-Bi LSTM模型的矿井涌水量预测[J].煤矿安全，2023,54(5):56-62.LIANG M,YIN S,YAO H,et al.Mine water inflow prediction based on DRN-Bi LSTM model[J].Safety in Coal Mines,2023,54(5):56-62.