计算机科学

计算机科学 ›› 2025, Vol. 52 ›› Issue (6A): 240400188-7.doi: 10.11896/jsjkx.240400188

• 大数据&数据科学 • 上一篇    下一篇

基于深度学习的矿井瓦斯浓度预测算法研究与实现

王宝会1, 高瞻1, 徐林2, 谭英洁1   

  1. 1 北京航空航天大学软件学院 北京 100191
    2 中国科学院空天信息创新研究院 北京 100094
  • 出版日期:2025-06-16 发布日期:2025-06-12
  • 通讯作者: 王宝会(wangbh@buaa.edu.cn)
  • 基金资助:
    矿井瓦斯防治大数据预警及可视化交互系统研究

Research and Implementation of Mine Gas Concentration Prediction Algorithm Based on Deep Learning

WANG Baohui1, GAO Zhan1, XU Lin2, TAN Yingjie1   

  1. 1 School of Software,Beihang University,Beijing 100191,China
    2 Aerospace Information Research Institute,Chinese Academy of Sciences,Beijing 100094,China
  • Online:2025-06-16 Published:2025-06-12
  • About author:WANG Baohui,born in 1973,senior engineer,master supervisor.His main research interests include software architecture,big data,artificial intelligence,etc.
  • Supported by:
    Big Data Warning and Visual Interactive System for Mine Gas Prevention and Control.

PDF (PC)

40

摘要： 目前国内外构建瓦斯浓度传统预测算法主要是ARIMA模型和SVM模型。随着深度学习技术的快速发展以及神经网络的兴起,最新的瓦斯浓度预测通过循环神经网络模型进行预测。循环神经网络因为具有非线性特点,并且考虑到了数据间的联系,所以预测效果相比传统预测算法有了进一步提升。而当样本序列长度加长时,由于其模型固有缺陷,预测能力会降低。文中针对此问题提出了一种新型的瓦斯浓度预测模型。卷积神经网络结合循环神经网络的方式,并且加入注意力机制增加数据间的表达能力。通过使用山西汾西矿业集团中兴煤业1209工作面的实际数据进行测试,传统的循环神经网络模型预测的平均相对误差为0.042 1,所提模型预测的平均相对误差为0.029 3。实验表明提出的算法相比瓦斯浓度传统预测算法获得了更好的预测性能。

关键词: 瓦斯浓度预测, 深度学习, 卷积神经网络, 循环神经网络, Attention机制, LSTM

Abstract: Currently,the traditional prediction algorithms for gas concentration both domestically and internationally primarily rely on ARIMA and SVM models.With the rapid development of deep learning technology and the rise of neural networks,the la-test gas concentration prediction is conducted through recurrent neural network(RNN) models.Due to their nonlinear characteris-tics and consideration of data connections,RNNs have further improved the prediction performance compared to traditional prediction algorithms.However,as the length of the sample sequence increases,the prediction ability decreases due to inherent flaws in the model.In response to this issue,the paper proposes a novel gas concentration prediction model.This model combines con-volutional neural networks (CNNs) with RNNs and incorporates an attention mechanism to enhance the expressive power between data.Through testing using actual data from the 1209 working face of Zhongxing Coal Industry in Shanxi Fenxi Mining Group,the average relative error predicted by the traditional RNN model is 0.42 1,while the average relative error predicted by the proposed model is 0.029 3.The experiment demonstrates that the proposed algorithm achieves better prediction performance compared to traditional gas concentration prediction algorithms.

Key words: Gas concentration prediction, Deep learning, Convolutional neural network, Recurrent neural network, Attention mechanism, LSTM

中图分类号: 

  • TP391
[1]LI D F,ZHAO S,WU F.A research ofcoal and gas outburst prediction based on ICA-SVM[J].Industry and Mine Automation,2009,35(10):36-38.
[2]LI D G,ZHAO S,YANG D P.A method of coal and gas outburst prediction based on KPCA-SVM[J].Industry and Mine Automation,2010,36(10):36-38.
[3]HUANG W Y.Research on mine gas early warning technology based on support vector machine data fusion[D].China University of Mining and Technology,2009.
[4]WU B,GUO Z G,WANG Z W.Prediction of Gas Emission in Mining Working Face Based on ARIMA-GM Model[J].Safty in Coal Mines,2015,46(11):152-155.
[5]FAN J D,HUANG Y X,YAN Z G,et al.Research on Gas Concentration Prediction Driven by ARIMA-SVM Combined Model[J].Industrial and Mining Automation,2022,48(9):134-139.
[6]QIAO M Y,MA X P,LAN J Y,et al.Short-term Gas Prediction Based on Weighted LS-SVM Time Series [J].Journal of Mining and Safety Engineering,2011,28(2):310-314.
[7]JIANG L.Construction and Simulation of Coal Mine Gas Concentration Prediction Model Based on BP Neural Network [J].Mining Safety and Environmental Protection,2010(4):37-39.
[8]JI Z G.Application of BP Neural Network Model Improved by Genetic Algorithm in Predicting Gas Emission from Adjacent Layers [J].Mine Safety,2011(7):36-38.
[9]GAO L,HU Y J,YU H Z.Gas Time Series Prediction Method Based on W_RBF [J].Journal of China Coal Society,2008,33(1):67-70.
[10]JIA P T,DENG J.Combined Prediction Model of Mine Gas Concentration Based on Generalized Average Operation [J].China Safety Science Journal,2012,22(6):41-46.
[11]SHAZEER N,MIRHOSEINI A,MAZIARZ K,et al.Outra-geously large neural networks:The sparsely-gated mixture-of-experts layer[J].arXiv:1701.06538,2017.
[12]LIU J Q.Research on Gas Data Time Series Prediction Based on Improved LSTM Recurrent Neural Network [D].China University of Mining and Technology,2019.
[13]XUN X Y,SU C,LI W,et al.Prediction of Coal Mine Gas Concentration Based on CNN-LSTM [J].Modern Information Technology,2020,4(20):149-152.
[14]QIN J X,GE S W,LONG F Q,et al.Prediction of Spatial-Temporal Distribution of Gas Concentration Based on GCN-GRU [J].Industrial and Mining Automation,2023,49(5):82-89,111.
[15]ZHOU H,ZHANG S,PENG J,et al.Informer:Beyond Efficient Transformer for Long Sequence Time-Series Forecasting[J].arXiv:2012.07436,2020.
[16]BRYAN L Ö,S A,NICOLAS L,et al.Temporal Fusion Transformers for interpretable multi-horizon time series forecasting[J].International Journal of Forecasting,2021,37(4):1748-1764.
[17]HU X,WANG W,TANG J,et al.Time series forecasting ofNOx concentration based on Informer for MSWI[C]//35th Chinese Control and Decision Conference(CCDC 2023).2023:319-324.
[18]OLEKSII K,BORIS G.Factorization tricks for LSTM networks[J].arXiv:1703.10722,2017.
[19]HOWARD A G,ZHU M,CHEN B,et al.Mobilenets:Efficient convolutional neural networks for mobile vision applications[J].arXiv:1704.04861,2017.
[20]KIM Y,DENTON C,HOANG L,et al.Structured attentionnetworks[C]//International Conference on Learning Representations.2017.
[21]LIN Z,FENG M W,NOGUEIRA D S C,et al.A structured self-attentive sentence embedding[J].arXiv:1703.03130,2017.
[22]GEHRING J,AULI M,GRANGIER D,et al.Dauphin.Convolutional sequence to sequence learning[J].arXiv:1705.03122v2,2017.
[23]ZENG A L,CHEN M X,ZHANG L,et al.Are transformers effective for time series forecasting？[C]//AAAI.2023.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发