卷积神经网络在储层预测中的应用研究

doi:10.11959/j.issn.1000-436x.2016240

摘要/Abstract

摘要：

人工神经网络作为人工智能的分支，在模式识别、分类预测等方面已成功地解决了许多现代计算机难以解决的实际问题。然而随着人工智能的发展，神经网络的自主性特征学习功能越来越重要，人工神经网络虽然表现出了良好的智能特性，但不能自主地学习特征。近年来，深度学习逐渐崛起，围绕深度神经网络的研究也越来越多，但其在地质储层参数预测领域的研究还很少。提出了一种应用卷积神经网络对地质储层参数进行预测的方法，该方法不仅能对储层参数进行精确预测，而且可以得到储层特征集。实验证明，卷积神经网络可以应用于地质储层参数预测，且预测精度较高，同时卷积神经网络的卷积特征为储层地质建模与测井资料解释提供了重要的支持。

关键词: 人工神经网络, 深度学习, 卷积神经网络, 储层预测, 储层特征

Abstract:

As the branch of artificial intelligence,artificial neural network solved many difficult practical problems in pattern recognition and classification prediction field successfully.However,they cannot learn the feature from networks.In recent years,deep learning becomes more and more advanced,but the research on the field of geological reservoir pa-rameter prediction is still rare.A method to predict reservoir parameters by convolutional neural network was presented,which can not only predict reservoir parameters accurately,but also get features of the geological reservoir.The study es-tablished the convolutional neural network model.Results show that the convolutional neural network can be used for reservoir parameter prediction,and get high prediction precision.Moreover,convolutional features from convolutional neural network provided important support for geological modeling and logging interpretation.

Key words: artificial neural network, deep learning, convolutional neural network, reservoir prediction, reservoir feature

段友祥,李根田,孙歧峰. 卷积神经网络在储层预测中的应用研究[J]. 通信学报, 2016, 37(Z1): 1-9.

You-xiang DUAN,Gen-tian LI,Qi-feng SUN. Research onconvolutional neural network for reservoir parameter prediction[J]. Journal on Communications, 2016, 37(Z1): 1-9.

图/表 12

图1

图2

图3

图4

表1

图5

图6

图7

图8

表2

图9

图10

参考文献 16

[1]	陈蓉，王峰 . 基于 MATLAB 的 BP 神经网络在储层物性预测中的应用[J]. 测井技术， 2009,33(1): 75-78. CHEN R , WANG F . Application of MATLAB-based of BP neural network in reservoir parameters prediction[J]. Well Logging Technol-ogy, 2009,33(1): 75-78.
[2]	HAMIDI H , RAFATI R . Prediction of oil reservoir porosity based on BP-ANN[C]// Innovation Management and Technology Re-search (ICIMTR), 2012 International Conference, 2012:241-246.
[3]	李映涛，袁晓宇，刘迪，等. BP神经网络在测井解释中的应用研究[J]. 西部探矿工程， 2013,25(3): 45-48. LI Y T , YUAN X Y , LIU D , et al. Research on the application of BP neural network in wells log interpretation[J]. West-china Exploration Engineering, 2013.
[4]	董兴朋 . 相似度-遗传神经网络在储层物性预测中的应用[J]. 测井技术， 2012,36(3): 267-271. DONG X P . Application of similarity-genetic neural network to reservoir parameters prediction[J]. Well Logging Technology, 2012.
[5]	潘少伟，梁鸿军，李良，等. 改进PSO-BP神经网络对储层参数的动态预测研究[J]. 计算机工程与应用， 2014,50(10): 52-56. PAN S , LIANG H , LI L , et al. Dynamic prediction on reservoir pa-rameter by improved PSO-BP neural network[J]. Computer Engineer-ing ＆Applications, 2014,50(10): 52-56.
[6]	BANESHI M , BEHZADIJO M , ROSTAMI M , et al. Using well logs to predict a multimin porosity model by optimized spread rbf net-works[J]. Energy Sources,Part A:Recovery,Utilization,and Envi-ronmental Effects, 2015,37(22): 2443-2450.
[7]	WU X J , JIANG G C , WANG X J , et al. Prediction of reservoir sensitivity using RBF neural network with trainable radial basis function[J]. Neural Computing ＆ Applications, 2013,22(5): 947-953.
[8]	NA'IMI S R , SHADIZADEH S R , RIAHI M A , et al. Estimation of reservoir porosity and water saturation based on seismic attributes us-ing support vector regression approach[J]. Journal of Applied Geo-physics, 2014,107:93-101.
[9]	AL-ANAZI A F , GATES I D . Support vector regression for porosity prediction in a heterogeneous reservoir:a comparative study[J]. Com-puters＆Geosciences, 2010,36(12): 1494-1503.
[10]	AL-ANAZI A F , GATES I D . Support vector regression to predict porosity and permeability:effect of sample size[J]. Computers ＆Geosciences, 2012,39(2): 64-76.
[11]	MOLLAHAN A . Estimation of reservoir water saturation using sup-port vector regression in an Iranian carbonate reservoir[C]// American Rock Mechanics Association. 2013.
[12]	KRIZHEVSKY A , SUTSKEVER I , HINTON G E . ImageNet classifi-cation with deep convolutional neural networks[J]. Advances in Neural Information Processing Systems, 2012,25(2): 2012.
[13]	DENG L , LI J , HUANG J T , et al. Recent advances in deep learning for speech research at microsoft[C]// 2013:8604-8608.
[14]	LECUN Y , BOTTOU L , BENGIO Y , et al. Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE, 1998,86(11):2278-2324.
[15]	李飞腾 . 卷积神经网络及其应用[D]. 大连理工大学， 2014. LI F T . Convolutional neural network and its applications[D]. Dalian University of Technology, 2014.
[16]	GLOROT X , BENGIO Y . Understanding the difficulty of training deep feedforward neural networks[J]. Journal of Machine Learning Research, 2010,9: 249-256.

序号	训练数据	预测数据	网络模型	描述
1	1号井，500组井深范围	1号井，1 000组井深范围	单层卷积
				2种卷积神经网络模型预测同井数据
2	(2 500~3 000 m)	(3 000~4 000 m)	双层卷积
3	1号井，1 000组井深范围	2号井，1 000组井深范围	单层卷积
				2种卷积神经网络模型预测邻井数据
4	(2 500~3 500 m)	(2 500~2 500 m)	双层卷积

序号	均方差(MSE)	相关系数(R)
1	0.000 721 08	0.870 0
2	0.000 604 38	0.890 7
3	0.000 794 26	0.946 0
4	0.000 761 23	0.942 8