Prediction of coalbed methane well productivity based on attention mechanism of CNN-GRU

ZHAO Haifeng; ZHU Likai; LIU Changsong; ZHANG Xianfan

doi:10.13347/j.cnki.mkaq.2023.12.004

Safety in Coal Mines > 2023 > 54(12): 11-17. > DOI: 10.13347/j.cnki.mkaq.2023.12.004

ZHAO Haifeng, ZHU Likai, LIU Changsong, ZHANG Xianfan. Prediction of coalbed methane well productivity based on attention mechanism of CNN-GRU[J]. Safety in Coal Mines, 2023, 54(12): 11-17. DOI: 10.13347/j.cnki.mkaq.2023.12.004

Citation:

PDF (1372 KB)

Prediction of coalbed methane well productivity based on attention mechanism of CNN-GRU

College of Petroleum Engineering, China University of Petroleum (Beijing), Beijing 102249, China

More Information

Received Date: August 18, 2022
Available Online: December 21, 2023

Graphical Abstract

Abstract

Abstract

In order to optimize the current prediction method of coalbed methane production and solve the problem of losing key information due to the influence of historical state of coalbed methane production data and too long sequence, a coalbed methane production prediction model based on attention mechanism of convolutional neural network was proposed. Firstly, the five drainage parameters with high correlation with daily gas production are selected, and the CBM productivity prediction model is constructed and trained based on the field drainage data of the well in Hancheng Block. Finally, the production capacity of the well area in the next 160 days is predicted by the model. The research results show that combined model CNN-GRUA with attention mechanism overcomes the problems that traditional prediction methods cannot deal with, including nonlinearity between data, time sequence and information loss, by extracting high-level features of data and learning the correlation of time series. Compared with BP neural network, convolutional neural network (CNN), gated recurrent unit(GRU) and CNN-GRU without attention mechanism model, the CNN-GRUA combined model has higher prediction accuracy, and the mean absolute percentage error is 1.72%.
- coalbed methane productivity prediction,
- deep learning,
- convolutional neural networks,
- gated recurrent units,
- attention mechanism

FullText(HTML)

References (19)

References

[1]	明盈,汤达祯,陶树,等. 煤层气水平井产能预测模型及其适用性研究[J]. 煤炭科学技术,2016,44(12):113−117. MING Ying, TANG Dazhen, TAO Shu, et al. Study on productivity prediction model of horizontal coalbed methane well and its applicability[J]. Coal Science and Technology, 2016, 44(12): 113−117.
[2]	任建华,任韶然,孟尚志. 煤层气井产气规律及产能影响因素分析[J]. 科学技术与工程,2013,13(10):2799−2802. doi: 10.3969/j.issn.1671-1815.2013.10.035 REN Jianhua, REN Shaoran, MENG Shangzhi. Analysis of production rule and productivity influencing factors of coalbed methane[J]. Science Technology and Engineering, 2013, 13(10): 2799−2802. doi: 10.3969/j.issn.1671-1815.2013.10.035
[3]	王衍,马俯波,李小军,等. 鄂东低煤阶U型井排采特征及产气量预测研究[J]. 中国煤炭地质,2018,30(8):29−35. WANG Yan, MA Fubo, LI Xiaojun, et al. Low ranked coal U-shaped CBM Well drainage features and production capacity prediction in Eastern Ordos Basin[J]. Coal Geology of China, 2018, 30(8): 29−35.
[4]	李亚林. 基于机器学习方法研究煤层气单井产量主控因素及产量预测[D]. 北京: 中国石油大学（北京）, 2017.
[5]	XIA T Q, GAO F, KANG J H, et al. A fully coupling coal–gas model associated with inertia and slip effectsfor CBM migration[J]. Environmental Earth Sciences, 2016, 75(7): 1−10.
[6]	SUN Z, SHI J T, WUKL, et al. A fully-coupled gas-water two phase productivity equations for low-permeability CBM wells[J]. Journal of Petroleum Science and Engineering, 2018, 166: 611−620. doi: 10.1016/j.petrol.2018.03.084
[7]	胡小虎,郑世毅,李保振,等. 物质平衡法对定容煤层气藏生产动态的预测[J]. 煤田地质与勘探,2011,39(3):29−32. doi: 10.3969/j.issn.1001-1986.2011.03.006 HU Xiaohu, ZHENG Shiyi, LI Baozhen, et al. Application of material balance method to deliverability prediction of closed coal-bed methane reservoir[J]. Coal Geology & Exploration, 2011, 39(3): 29−32. doi: 10.3969/j.issn.1001-1986.2011.03.006
[8]	杜严飞,吴财芳,薛佳佳,等. 基于人工神经网络煤层气井短期产能预测[J]. 辽宁工程技术大学学报(自然科学版),2013,32(4):493−498. DU Yanfei, WU Caifang, XUE Jiajia, et al. Coalbed methane well short-term productivity prediction by using artificial neural network[J]. Journal of Liaoning Technical University(Natural Science), 2013, 32(4): 493−498.
[9]	朱庆忠,胡秋嘉,杜海为,等. 基于随机森林算法的煤层气直井产气量模型[J]. 煤炭学报,2020,45(8):2846−2855. ZHU Qingzhong, HU Qiujia, DU Haiwei, et al. A gas production model of vertical coalbed methane well based on random forest algorithm[J]. Journal of China Coal Society, 2020, 45(8): 2846−2855.
[10]	董维强,孟召平,沈振,等. 基于循环神经网络的煤层气井产气量预测方法研究[J]. 煤炭科学技术,2021,49(9):176−183. DONG Weiqiang, MENG Zhaoping, SHEN Zhen, et al. Research on coalbed methane well gas production forecast method based on cyclic neural network[J]. Coal Science and Technology, 2021, 49(9): 176−183.
[11]	吕玉民,汤达祯,李治平,等. 煤层气井动态产能拟合与预测模型[J]. 煤炭学报,2011,36(9):1481−1485. LYU Yumin, TANG Dazhen, LI Zhiping, et al. Fitting and predicting models for coalbed methane wells dynamic productivity[J]. Journal of China Coal Society, 2011, 36(9): 1481−1485.
[12]	李艳芳,程建远,王成. 基于支持向量机的地震属性优选及煤层气预测[J]. 煤田地质与勘探,2012,40(6):75−78. LI Yanfang, CHENG Jianyuan, WANG Cheng. Seismic attribute optimization based on support vector machine and coalbed methane prediction[J]. Coal Geology & Exploration, 2012, 40(6): 75−78.
[13]	韩永成, 张闻文, 何伟基, 等. 基于自适应截断模拟曝光和无监督融合的低照度真彩色图像增强算法[J]. 光子学报, 2023, 52（9）: 237−251. HAN Yongcheng, ZHANG Wenwen, HE Weiji, et al. Low-light true color image enhancement algorithm based on adaptive truncation simulation exposure and unsupervised fusion[J]. Acta Photonica Sinica, 2023, 52（9）: 237−251.
[14]	JIAO R, HUANG X, MA X, et al. A model combiningstacked auto encoder and back propagation algorithmfor short-term wind power forecasting[J]. IEEE Access, 2018, 6: 17851−17858. doi: 10.1109/ACCESS.2018.2818108
[15]	杨涛,贺萌萌. 注意力分配对制造企业数字化转型的影响-基于fsQCA和NCA方法的研究[J]. 工业技术经济,2023,42(9):12−21. doi: 10.3969/j.issn.1004-910X.2023.09.002 YANG Tao, HE Mengmeng. The impact of attention allocation on digital transformation of manufacturing enterprises−research based on fsQCA and NCA methods[J]. Journal of Industrial Technological Economics, 2023, 42(9): 12−21. doi: 10.3969/j.issn.1004-910X.2023.09.002
[16]	BREIMAN L. Random forests[J]. Machine learning, 2001, 45(1): 5−32. doi: 10.1023/A:1010933404324
[17]	毛伊敏, 李文豪. 基于互信息和融合加权的并行深度森林算法[J/OL]. 计算机应用研究, 2023: 1−11.[2023-10-19]. DOI: 10.19734/j.issn.1001-3695.2023.05.0240. MAO Yimin, LI Wenhao. Parallel deep forest algorithm based on mutual information and mixed weighting[J/OL]. Application Research of Computers, 2023: 1−11.[2023-10-19]. DOI: 10.19734/j.issn.1001-3695.2023.05.0240.
[18]	伊伟,涂志民,冯延青,等. 韩城矿区煤层含气性分布规律及地质控制因素研究[J]. 煤炭科学技术,2017,45(7):156−160. YI Wei, TU Zhimin, FENG Yanqing, et al. Study on distribution laws of gas-bearing property of coal seams and geological control factors in Hancheng Mining Area[J]. Coal Science and Technology, 2017, 45(7): 156−160.
[19]	HU Jinhui, XIA Dawen, CHENG Huqiang, et al. A decentralized Nesterov gradient method for stochastic optimization over unbalanced directed networks[J]. Asian Journal of Control, 2020, 24(2): 576−593.

[1]	XU Jiang, CHEN Qian, LI Juyun. Molecular simulation of self-diffusion of CH₄, H₂O and CO₂ in coal macromolecule[J]. Safety in Coal Mines, 2024, 55(9): 85-93. DOI: 10.13347/j.cnki.mkaq.20221827
[2]	DENG Xiaopeng, XIANG Jianhua. Molecular simulation study on competitive adsorption characteristics of CO₂ and CH₄ for 8^# coal in Dongqu Mine[J]. Safety in Coal Mines, 2024, 55(3): 18-24. DOI: 10.13347/j.cnki.mkaq.20230796
[3]	WANG Wencai, WANG Peng, CAO Zhao, LI Junpeng. Simulation of gas molecules permeation diffusion in coal structure and gangue slit structure[J]. Safety in Coal Mines, 2023, 54(11): 33-41. DOI: 10.13347/j.cnki.mkaq.2023.11.007
[4]	GAO Yu. Study on mathematical model of coal particles gas adsorption under different particle sizes and pressures[J]. Safety in Coal Mines, 2023, 54(7): 109-117.
[5]	LI Ziwen, BAI Yansong, YU Hongjin, GAO Yabin. Molecular simulation of adsorption thermodynamic properties of CH₄ and CO₂ at different temperature and moisture content conditions[J]. Safety in Coal Mines, 2022, 53(10): 112-119.
[6]	ZHANG Hongtu, HAO Yushuang, WEI Jianping. Study on Characteristic Parameters of Particulate Coal Gas Desorption and Diffusion Under Negative Pressure Environment[J]. Safety in Coal Mines, 2020, 51(10): 191-195.
[7]	XU Chao, WANG Shuo, GUO Haijun, FU Qiang. Unsteady-state Diffusion Characteristics Model of Gas in Coal and Experimental Verification[J]. Safety in Coal Mines, 2018, 49(5): 1-5.
[8]	LIU Yanwei, LI Tong, XUE Wentao. Difference of Dynamic Variation Laws of Gas Diffusion Coefficient of Soft and Hard Coal[J]. Safety in Coal Mines, 2017, 48(9): 17-20,24.
[9]	CHI Lei-lei, WANG Qi-fei, WANG Fei-yin, LI Zhen, DONG Li-hui, SHI Quan-shou. Coal Gas Desorption and Diffusion Laws and Its Experimental Study[J]. Safety in Coal Mines, 2013, 44(12): 1-3,10.
[10]	LI Yi-bo, ZHENG Wan-cheng, WANG Feng-shuang. The Effect of Coal Sample Particle Size on Coal Adsorption Constants and Initial Speed of Methane Diffusing[J]. Safety in Coal Mines, 2013, 44(1): 5-8.

Cited By

Cited by

Periodical cited type(3)

1.	贾丽娜. 基于散斑干涉法的岩体剪切变形程度研究——以铜仁市水库工程为例. 水利科技与经济. 2024(08): 6-9 .
2.	李明胜. 基于数值模拟分析的煤场软土地基处理. 中国煤炭地质. 2024(11): 44-49+39 .
3.	陈光波，唐薇，李谭，王创业，王二雨，张国华. 裂隙煤岩组合体单轴压缩力学响应及失稳机制. 岩土力学. 2024(09): 2633-2652 .

Other cited types(2)

Get Citation

PDF

XML

Article views (78) PDF downloads (24) Cited by(5)

Turn off MathJax

Article Contents

Abstract

References

Prediction of coalbed methane well productivity based on attention mechanism of CNN-GRU