石油勘探与开发 >

2020 , Vol. 47 >Issue 5: 1009 - 1015

DOI: https://doi.org/10.11698/PED.2020.05.15

油气田开发

基于循环神经网络的油田特高含水期产量预测方法

  • 王洪亮 ,
  • 穆龙新 ,
  • 时付更 ,
  • 窦宏恩
展开
  • 中国石油勘探开发研究院,北京 100083
王洪亮(1984-),男,黑龙江哈尔滨人,中国石油勘探开发研究院在读博士研究生,主要从事统计建模、油气领域大数据与人工智能研究。地址:北京市海淀区学院路20号,中国石油勘探开发研究院计算机应用技术研究所,邮政编码:100083。E-mail:whldqpi@126.com

收稿日期: 2019-09-20

  修回日期: 2020-06-29

  网络出版日期: 2020-09-22

基金资助

国家科技重大专项“大型油气田及煤层气开发”(2016ZX05016-006)

收起

Production prediction at ultra-high water cut stage via Recurrent Neural Network

  • WANG Hongliang ,
  • MU Longxin ,
  • SHI Fugeng ,
  • DOU Hongen
Expand
  • PetroChina Research Institute of Petroleum Exploration & Development, Beijing 100083, China

Received date: 2019-09-20

  Revised date: 2020-06-29

  Online published: 2020-09-22

Fold

摘要

根据油田生产历史数据利用深度学习方法预测油田特高含水期产量,并进行了实验验证和应用效果分析。考虑到传统全连接神经网络(FCNN)无法描述时间序列数据的相关性,基于一种循环神经网络(RNN)即长短期记忆神经网络(LSTM)来构建油田产量预测模型。该模型不仅考虑了产量指标与其影响因素之间的联系,还兼顾了产量随时间变化的趋势和前后关联。利用国内某中高渗透砂岩水驱开发油田生产历史数据进行特高含水期产量预测,并与传统水驱曲线方法和FCNN的预测结果比较,发现基于深度学习的LSTM预测精度更高,针对油田生产中复杂时间序列的预测结果更准确。利用LSTM模型预测了另外两个油田的月产油量,预测结果较好,验证了方法的通用性。图3表3参40

关键词： 产量预测; 特高含水期; 机器学习; 长短期记忆神经网络; 人工智能

本文引用格式

王洪亮 , 穆龙新 , 时付更 , 窦宏恩 . 基于循环神经网络的油田特高含水期产量预测方法[J]. 石油勘探与开发, 2020 , 47(5) : 1009 -1015 . DOI: 10.11698/PED.2020.05.15

Abstract

A deep learning method for predicting oil field production at ultra-high water cut stage from the existing oil field production data was presented, and the experimental verification and application effect analysis were carried out. Since the traditional Fully Connected Neural Network (FCNN) is incapable of preserving the correlation of time series data, the Long Short-Term Memory (LSTM) network, which is a kind of Recurrent Neural Network (RNN), was utilized to establish a model for oil field production prediction. By this model, oil field production can be predicted from the relationship between oil production index and its influencing factors and the trend and correlation of oil production over time. Production data of a medium and high permeability sandstone oilfield in China developed by water flooding was used to predict its production at ultra-high water cut stage, and the results were compared with the results from the traditional FCNN and water drive characteristic curves. The LSTM based on deep learning has higher precision, and gives more accurate production prediction for complex time series in oil field production. The LSTM model was used to predict the monthly oil production of another two oil fields. The prediction results are good, which verifies the versatility of the method.

Key words： production prediction; ultra-high water cut; machine learning; Long Short-Term Memory; artificial intelligence

参考文献

[1] 黄广庆. 特高含水期产量递减分析及递减率表征公式[J]. 科学技术与工程, 2019, 19(15): 99-104.
HUANG Guangqing.Production decline analysis and characterization formula of decline rate at the ultra-high water cut stage[J]. Science Technology and Engineering, 2019, 19(15): 99-104.
[2] 王继强, 石成方, 纪淑红, 等. 特高含水期新型水驱特征曲线[J]. 石油勘探与开发, 2017, 44(6): 955-960.
WANG Jiqiang, SHI Chengfang, JI Shuhong, et al.New water drive characteristic curves at ultra-high water cut stage[J]. Petroleum Exploration and Development, 2017, 44(6): 955-960.
[3] 陈元千, 陶自强. 高含水期水驱曲线的推导及上翘问题的分析[J].断块油气田, 1997, 4(3): 19-24.
CHEN Yuanqian, TAO Ziqiang.Derivation of water drive curve at high water-cut stage and its analysis of upwarding problem[J]. Fault-Block Oil & Gas Field, 1997, 4(3): 19-24.
[4] 刘晓华, 邹春梅, 姜艳东, 等. 现代产量递减分析基本原理与应用[J]. 天然气工业, 2010, 30(5): 50-54.
LIU Xiaohua, ZOU Chunmei, JIANG Yandong, et al.Basic principle and application of modern production decline analysis[J]. Natural Gas Industry, 2010, 30(5): 50-54.
[5] 张倩倩. 产量递减分析方法简评[J]. 油气地球物理, 2013, 11(3): 41-44.
ZHANG Qianqian.A brief review of production decline analysis methods[J]. Oil and Gas Geophysics, 2013, 11(3): 41-44.
[6] CRNKOVIC-FRIIS L, ERLANDSON M.Geology driven EUR prediction using deep learning[R]. SPE 174799-MS, 2015.
[7] GU M, GOKARAJU D, CHEN D, et al.Shale fracturing characterization and optimization by using anisotropic acoustic interpretation, 3D fracture modeling, and supervised machine learning[J]. Petrophysics, 2016, 57(6): 573-587.
[8] SIDAHMED M, ROY A, SAYED A.Streamline rock facies classification with deep learning cognitive process[R]. SPE 187436-MS, 2017.
[9] WU P, JAIN V, KULKARNI M S, et al.Machine learning-based method for automated well-log processing and interpretation[M]//ALUMBAUGH D, BEVC D. SEG technical program expanded abstracts 2018. Tulsa: Society of Exploration Geophysicists, 2018: 2041-2045.
[10] NOSHI C I, ASSEM A I, SCHUBERT J J.The role of big data analytics in exploration and production: A review of benefits and applications[R]. SPE 193776-MS, 2018.
[11] PHAM N, FOMEL S, DUNLAP D.Automatic channel detection using deep learning[M]//ALUMBAUGH D, BEVC D. SEG technical program expanded abstracts 2018. Tulsa: Society of Exploration Geophysicists, 2018: 2026-2030.
[12] LI W.Classifying geological structure elements from seismic images using deep learning[M]//ALUMBAUGH D, BEVC D. SEG technical program expanded abstracts 2018. Tulsa: Society of Exploration Geophysicists, 2018: 4643-4648.
[13] MEHTA A.Tapping the value from big data analytics[J]. Journal of Petroleum Technology, 2016, 68(12): 40-41.
[14] HALL B.Facies classification using machine learning[J]. The Leading Edge, 2016, 35(10): 906-909.
[15] CARPENTER C.Geology-driven estimated-ultimate-recovery prediction with deep learning[J]. Journal of Petroleum Technology, 2016, 68(5): 74-75.
[16] MA S M.Technology focus: Formation evaluation (August 2018)[J]. Journal of Petroleum Technology, 2018, 70(8): 50.
[17] SAPUTELLI L.Technology focus: Petroleum data analytics[J]. Journal of Petroleum Technology, 2016, 68(10): 66.
[18] LI H, MISRA S.Long short-term memory and variational autoencoder with convolutional neural networks for generating NMR T2 distributions[J]. IEEE Geoscience and Remote Sensing Letters, 2018, 16(2): 192-195.
[19] JOBE T D, VITAL-BRAZIL E, KHAIF M.Geological feature prediction using image-based machine learning[J]. Petrophysics, 2018, 59(6): 750-760.
[20] SILVA A A, LIMA NETO I A, MISSÁGIA R M, et al.Artificial neural networks to support petrographic classification of carbonate-siliciclastic rocks using well logs and textural information[J]. Journal of Applied Geophysics, 2015, 117: 118-125.
[21] 柴艳军. 基于灰色关联法的页岩气水平井产量主控因素分析[J]. 重庆科技学院学报(自然科学版), 2018, 20(2): 32-34.
CHAI Yanjun.Analysis of main controlling factors of shale gas horizontal well production based on grey correlation method[J]. Journal of Chongqing University of Science and Technology (Natural Science Edition), 2018, 20(2): 32-34.
[22] 王忠东, 王业博, 董红, 等. 页岩气水平井产量主控因素分析及产能预测[J]. 测井技术, 2017, 41(5): 577-582.
WANG Zhongdong, WANG Yebo, DONG Hong, et al.Analysis of production control factors and productivity prediction of horizontal wells in shale gas[J]. Journal of Well Logging Technology, 2017, 41(5): 577-582.
[23] 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.
LI Yalin.Study on main control factors and production prediction of single well production of coalbed methane based on machine learning[D]. Beijing: China University of Petroleum (Beijing), 2017.
[24] 李春生, 谭民浠, 张可佳. 基于改进型BP神经网络的油井产量预测研究[J]. 科学技术与工程, 2011, 11(31): 7766-7769.
LI Chunsheng, TAN Minxi, ZHANG Kejia.Oil well production prediction based on improved BP neural network[J]. Science and Technology and Engineering, 2011, 11(31): 7766-7769.
[25] 田亚鹏, 鞠斌山. 基于遗传算法改进BP神经网络的页岩气产量递减预测模型[J]. 中国科技论文, 2016, 11(15): 1710-1715.
TIAN Yapeng, JU Binshan.Shale gas production decline prediction model based on improved BP neural network based on genetic algorithm[J]. Chinese Science and Technology Paper, 2016, 11(15): 1710-1715.
[26] 马林茂, 李德富, 郭海湘, 等. 基于遗传算法优化BP神经网络在原油产量预测中的应用: 以大庆油田BED试验区为例[J]. 数学的实践与认识, 2015, 45(24): 117-128.
MA Linmao, LI Defu, GUO Haixiang, et al.Application of BP neural network optimization based on genetic algorithm in crude oil production prediction: Take BED test area of Daqing Oilfield as an example[J]. Practice and Understanding of Mathematics, 2015, 45(24): 117-128.
[27] 樊灵, 赵孟孟, 殷川, 等. 基于BP神经网络的油田生产动态分析方法[J]. 断块油气田, 2013, 20(2): 204-206.
FAN Ling, ZHAO Mengmeng, YIN Chuan, et al.Dynamic analysis method of oil field production based on BP neural network[J]. Fault Block Oil and Gas Field, 2013, 20(2): 204-206.
[28] 杨婷婷. 基于人工神经网络的油田开发指标预测模型及算法研究[D]. 大庆: 东北石油大学, 2013.
YANG Tingting.Research on prediction model and algorithm of oilfield development index based on artificial neural network[D]. Daqing: Northeast Petroleum University, 2013.
[29] 谷建伟, 隋顾磊, 李志涛, 等. 基于ARIMA-Kalman滤波器数据挖掘模型的油井产量预测[J]. 深圳大学学报(理工版), 2018, 35(6): 575-581.
GU Jianwei, SUI Gulei, LI Zhitao, et al.Oil well production prediction based on ARIMA-Kalman filter data mining model[J]. Journal of Shenzhen University (Science and Technology Edition), 2018, 35(6): 575-581.
[30] 李达. 基于时间序列分析方法的油田产量预测与应用[D]. 兰州: 兰州理工大学, 2018.
LI Da.Oilfield production prediction and application based on time series analysis method[D]. Lanzhou: Lanzhou University of Technology, 2018.
[31] HOCHREITER S, SCHMIDHUBER J.Long short-term memory[J]. Neural Computation, 1997, 9(8): 1735-1780.
[32] 谷建伟, 周梅, 李志涛, 等. 基于数据挖掘的长短期记忆网络模型油井产量预测方法[J]. 特种油气藏, 2019, 26(2): 77-81.
GU Jianwei, ZHOU Mei, LI Zhitao, et al.Prediction method of oil well production based on data mining based on long-term and short-term memory network model[J]. Special Oil and Gas Reservoirs, 2019, 26(2): 77-81.
[33] 侯春华. 基于长短期记忆神经网络的油田新井产油量预测方法[J]. 油气地质与采收率, 2019, 26(3): 105-110.
HOU Chunhua.Prediction method of oil production in new wells based on long and short term memory neural network[J]. Oil and Gas Geology and Recovery, 2019, 26(3): 105-110.
[34] MIKOLOV T, KARAFIÁT M, BURGET L, et al.Recurrent neural network based language model[R]. Makuhari, Chiba, Japan: 11th Annual Conference of the International Speech Communication Association, 2010.
[35] GRAVES A, JAITLY N. Towards end-to-end speech recognition with recurrent neural networks[C]//Proceedings of the 31st International Conference on Machine Learning. Washington D.C.: IEEE Computer Society Press, 2014: 1764-1772.
[36] SUTSKEVER I, VINYALS O, LE Q V.Sequence to sequence learning with neural networks[R]. Montreal, Canada: Advances in Neural Information Processing Systems 27 (NIPS 2014), 2014.
[37] 杨祎玥, 伏潜, 万定生. 基于深度循环神经网络的时间序列预测模型[J]. 计算机技术与发展, 2017, 27(3): 35-38.
YANG Yiyue, FU Qian, WAN Dingsheng.A prediction model for time series based on deep recurrent neural network[J]. Computer Technology and Development, 2017, 27(3): 35-38.
[38] 黄婷婷, 余磊. SDAE-LSTM模型在金融时间序列预测中的应用[J]. 计算机工程与应用, 2019, 55(1): 142-148.
HUANG Tingting, YU Lei.Application of SDAE-LSTM model in financial time series forecasting[J]. Computer Engineering and Application, 2019, 55(1): 142-148.
[39] DAS S, GILES C L, SUN G.Learning context-free grammars: Capabilities and limitations of a recurrent neural network with an external stack memory[C]//Proceedings of the Fourteenth Annual Conference of the Cognitive Science Society. New York: Cognitive Science Society, 1992.
[40] GUYON I, WESTON J, BARNHILL S, et al.Gene selection for cancer classification using support vector machines[J]. Machine Learning, 2002, 46(1/2/3): 389-422.
Options
文章导航

/

〈 〉

版权所有：《石油勘探与开发》编辑部；《石油勘探与开发（英文）》编辑部
地址：北京市海淀区学院路20号，中国石油勘探开发研究院数据中心605室　

京ICP备15044687号    技术支持：北京玛格泰克科技发展有限公司