吉林大学学报(地球科学版) ›› 2020, Vol. 50 ›› Issue (1): 208-216.doi: 10.13278/j.cnki.jjuese.20190055

• 地质工程与环境工程 • 上一篇    

基于多变量LSTM神经网络的地下水水位预测

闫佰忠1,2,3, 孙剑1,2,3, 王昕洲4, 韩娜1,2,3, 刘博5   

  1. 1. 河北地质大学水资源与环境学院, 石家庄 050031;
    2. 河北省水资源可持续利用与开发重点实验室, 石家庄 050031;
    3. 河北省水资源可持续利用与产业结构化协同创新中心, 石家庄 050031;
    4. 河北省地质资源环境监测与保护重点实验室(筹), 石家庄 050031;
    5. 沈阳环境科学研究院, 沈阳 110167
  • 收稿日期:2019-03-23 发布日期:2020-02-11
  • 作者简介:闫佰忠(1988-),男,副教授,博士,主要从事水文地质、地热资源等方面的研究,E-mail:jluybz@126.com
  • 基金资助:
    中国博士后基金面上项目(2018M631874);河北省自然科学基金项目(D2018403040);河北省高等学校科学技术研究项目(ZD2019082);河北地质大学博士科研启动基金项目(BQ2017011);河北省水利科技计划项目(2017-59);河北省地质资源环境监测与保护重点实验室开放基金项目(JCYKT201901);河北省地矿局科研项目(454-0601-YBN-U1MR)

Multivariable LSTM Neural Network Model for Groundwater Levels Prediction

Yan Baizhong1,2,3, Sun Jian1,2,3, Wang Xinzhou4, Han Na1,2,3, Liu Bo5   

  1. 1. School of Water Resources&Environment, Hebei GEO University, Shijiazhuang 050031, China;
    2. Key Laboratory of Sustained Utilization&Development of Water Resources, Hebei Province, Shijiazhuang 050031, China;
    3. Collaborative Innovation Center for Sustainable Utilization of Water Resources and Optimization of Industrial Structure, Hebei Province, Shijiazhuang 050031, China;
    4. Key Laboratory of Geological Resources and Environment Monitoring and Protection, Hebei Province, Shijiazhuang 050031, China;
    5. Shenyang Academy of Environmental Sciences, Shenyang 110167, China
  • Received:2019-03-23 Published:2020-02-11
  • Supported by:
    Supported by China's Post-Doctoral Science Fund (2018M631874), Natural Science Fund Project in Hebei Province (D2018403040), Scientific Research Projects of the Higher University in Hebei (ZD2019082), Scientific Research Initiation Funds for PhD Scholars of GEO University (BQ2017011), Water Conservancy Science and Technology Plan Projects of Hebei Province (2017-59), Key Laboratory of Geological Resources and Environmental Monitoring and Protection Fund of Hebei Province (JCYKT201901) and Scientific Research Project of Hebei Bureau of Geology and Mineral Resources (454-0601-YBN-U1MR)

摘要: 为解决以往模型未考虑地下水位相关影响因素的问题,探讨长短期记忆(LSTM)神经网络在地下水位预测中的应用,利用长短期记忆神经网络,采用多变量输入的方式,构建了基于多变量LSTM神经网络的地下水水位预测模型。以泰安市岱岳区J1号监测井为例,采用2001-2014年地下水水位动态监测资料与相关影响因素数据,利用多变量LSTM神经网络对2015-2016年地下水位进行预测,并与单变量LSTM神经网络和反向传播(BP)神经网络进行对比。研究结果表明:以相关影响变量为输入的BP神经网络无法考虑时序变化规律,预测均方根误差最大,为2.399 3;以地下水位为变量输入的单变量LSTM神经网络仅能根据时序变化作出相应预测,无法考虑相关变量影响,预测均方根误差为2.102 2;基于多变量输入的LSTM神经网络的预测精度显著高于单变量LSTM神经网络和BP神经网络,预测均方根误差最小,仅为1.919 1。总体上,多变量LSTM神经网络地下水位预测模型仅在某些峰值处误差较大,但总体预测效果较为理想。

关键词: 地下水位预测, LSTM, 多变量

Abstract: The effect of the traditional groundwater level prediction model is not ideal due to the lack of considering the temporal and spatial variation rules and the relevant influencing factors of groundwater level.The authors used the long and short term memory neural network (LSTM), adopted the method of multivariable input,and constructed the groundwater level prediction model. Taking the monitoring Well J1 in Daiyue district of Tai'an City as an example, the groundwater level dynamic monitoring data from 2001 to 2014 and the relevant influencing factor data were used to predict the groundwater level from 2015 to 2016 by using the multivariable LSTM neural network, and further compared with the single variable LSTM neural network model and BP neural network model. The research results show that the BP neural network prediction model can only make the corresponding prediction according to the change of the influencing variables without considering the change rule of timing sequence, so the prediction error is large (2.399 3). The single-variable LSTM model only takes the groundwater level as the variable input with no considering the influence of relevant factors, and the prediction error is 2.102 2. The prediction accuracy of the LSTM neural network prediction model based on multivariable input is significantly higher than that of the univariate LSTM neural network and BP neural network models, and the prediction root mean square error is only 1.919 1, which successfully verifies the accuracy of the multivariable LSTM neural network groundwater level prediction model.

Key words: groundwater level prediction, long short-term memory, multivariate

中图分类号: 

  • P641.8
[1] 刘博,肖长来,梁秀娟. SOM-RBF神经网络模型在地下水位预测中的应用[J]. 吉林大学学报(地球科学版),2015,45(1):225-231. Liu Bo, Xiao Changlai, Liang Xiujuan. Application of Combining SOM and RBF Neural Network Model for Groundwater Levels Prediction[J]. Journal of Jilin University (Earth Science Edition),2015,45(1):225-231.
[2] 马新双,卢胜勇,李明良,等. 北方浅层地下水超采区地下水预测模型研究[J]. 南水北调与水利科技,2011,9(4):134-136. Ma Xinshuang, Lu Shengyong, Li Mingliang, et al. Research of Groundwater Prediction Model in Shallow Groundwater Over-Exploitation Zone in the Northern Area[J]. South-to-North Water Transfers and Water Science & Technology,2011,9(4):134-136.
[3] 周振民,赵明亮,李玲. GM(1,1)模型在滦河下游地区地下水位预测中的应用[J]. 中国农村水利水电,2011(2):50-52. Zhou Zhenmin, Zhao Mingliang, Li Ling. The Application of GM (1, 1) Model to the Prediction of Underground Water Level in the Downstream of the Luanhe River[J]. China Rural Water and Hydropower,2011(2):50-52.
[4] Ahn H. Modeling of Groundwater Heads Based on Second-Order Difference Time Series Models[J]. Journal of Hydrology, 2000, 234(Sup.1/2):82-94.
[5] 陈汉军,杨雪,黄东卫. 时间序列在地下水位预测中的应用[J]. 天津理工大学学报,2008,24(2):8-10. Chen Hanjun, Yang Xue, Huang Dongwei. Application of Time Series to Groundwater Level Prediction[J]. Journal of Tianjin University of Technology,2008,24(2):8-10.
[6] 张展羽,梁振华,冯宝平,等. 基于主成分-时间序列模型的地下水位预测[J]. 水科学进展,2017(3):97-102. Zhang Zhanyu, Liang Zhenhua, Feng Baoping, et al. Groundwater Level Forecast Based on Principal Component Analysis and Multivariate Time Series Model[J]. Advances in Water Science,2017(3):97-102.
[7] Yang Y, Dong J, Sun X, et al. A CFCC-LSTM Model for Sea Surface Temperature Prediction[J]. IEEE Geoscience and Remote Sensing Letters, 2017, 15(2):1-5.
[8] Tax N,Verenich I, La Rosa M, et al. Predictive Business Process Monitoring with LSTM Neural Networks[M]. Advanced Information Systems Engineering, 2017(5):477-492.
[9] Ma X, Tao Z, Wang Y, et al.Long Short-Term Memory Neural Network for Traffic Speed Prediction Using Remote Microwave Sensor Data[J]. Transportation Research:Part C:Emerging Technologies, 2015, 54:187-197.
[10] Hochreiter S, Schmidhuber J. Long Short-Term Memory[J]. Neural Computation, 1997, 9(8):1735-1780.
[11] Greff K, Srivastava R K, Koutník Jan, et al. LSTM:A Search Space Odyssey[J]. IEEE Transactions on Neural Networks & Learning Systems, 2015, 28(10):2222-2232.
[12] Newey W K. Adaptive Estimation of Regression Models via Moment Restrictions[J]. Journal of Econometrics, 1988, 38(3):301-339.
[13] Kingma D P, Ba J. Adam:A Method for Stochastic Optimization[J]. Computer Science, 2014,32(3):111-125.
[14] 祁晓凡,李文鹏,李海涛, 等. 济南岩溶泉域地下水位、降水、气温与大尺度气象模式的遥相关[J]. 水文地质工程地质,2015, 42(6):18-28. Qi Xiaofan, Li Wenpeng, Li Haitao, et al. Teleconnections Between Groundwater Levels, Precipitation, Air Temperature of the Jinan Karst Springs Watershed and Large Scale Climatic Patterns[J]. Hydrogeology & Engineering Geology,2015, 42(6):18-28.
[15] 杨建中. 地下水动态的影响因素分析[J]. 图书情报导刊, 2008, 18(9):121-122. Yang Jianzhong. Analysis on the Factors Influencing the Groundwater Regime[J]. Journal of Library and Information Science,2008, 18(9):121-122.
[16] 王鑫,吴际,刘超,等. 基于LSTM循环神经网络的故障时间序列预测[J]. 北京航空航天大学学报,2018,44(4):772-784. Wang Xin, Wu Ji, Liu Chao,et al. Exploring LSTM Based Recurrent Neural Network for Failure Time Series Prediction[J]. Journal of Beijing University of Aeronautics and Astronautics,2018,44(4):772-784.
[17] Ciftcioglu O, Turkcan E. Selection of Hidden Layer Nodes in Neural Networks by Statistical Tests[J]. Macromolecules, 1992, 40(17):6217-6223.
[18] 管新建,逯洪波,徐清山. 基于BP神经网络的区域地下水位动态预测[J]. 人民黄河,2006,28(8):40-41. Guan Xinjian, Lu Hongbo, Xu Qingshan. Dynamic Prediction of Regional Groundwater Level based on BP Neural Network[J]. Yellow River,2006,28(8):40-41.
[1] 刘博, 肖长来, 梁秀娟. SOM-RBF神经网络模型在地下水位预测中的应用应用[J]. 吉林大学学报(地球科学版), 2015, 45(1): 225-231.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!