为了在不增加经济成本的基础上补充缺失的测井信息,提出利用机器学习方法根据已有的部分测井曲线生成人工测井曲线,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）无法描述数据的空间相关性,基于一种循环神经网络（RNN）即长短期记忆神经网络（LSTM）来构建测井曲线生成方法。该方法生成的曲线不仅考虑了不同测井曲线的内在联系,同时兼顾了测井信息随深度的变化趋势和前后关联。将标准LSTM与串级系统相结合,提出了一种串级长短期记忆神经网络（CLSTM）。采用真实测井数据进行实验,LSTM明显优于传统FCNN,生成的测井数据精度更高;CLSTM更适用于测井曲线生成这种多序列数据问题;提出的基于机器学习的人工测井曲线生成方法更准确经济。图8表2参32
To supplement missing logging information without increasing economic cost, a machine learning method to generate synthetic well logs from the existing log 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 spatial dependency, the Long Short-Term Memory (LSTM) network, which is a kind of Recurrent Neural Network (RNN), was utilized to establish a method for log reconstruction. By this method, synthetic logs can be generated from series of input log data with consideration of variation trend and context information with depth. Besides, a cascaded LSTM was proposed by combining the standard LSTM with a cascade system. Testing through real well log data shows that: the results from the LSTM are of higher accuracy than the traditional FCNN; the cascaded LSTM is more suitable for the problem with multiple series data; the machine learning method proposed provides an accurate and cost effective way for synthetic well log generation.
[1] ASQUITH G B, KRYGOWSKI D, GIBSON C R.Basic well log analysis[M]. Tulsa: American Association of Petroleum Geologists, 2004.
[2] BATEMAN R M.Openhole log analysis and formation evaluation[M]. 2nd ed. Richardson, Texas: Society of Petroleum Engineers, 2012
[3] ESKANDARI H, REZAEE M, MOHAMMADNIA M.Application of multiple regression and artificial neural network techniques to predict shear wave velocity from wireline log data for a carbonate reservoir South-West Iran[J]. CSEG Recorder, 2004, 29(7): 40-48.
[4] WENDT W, SAKURAI S T, NELSON P.Permeability prediction from well logs using multiple regression[M]. Amsterdam: Elsevier, 1986: 181-221.
[5] ZERROUKI A A, AIFA T, BADDARI K.Prediction of natural fracture porosity from well log data by means of fuzzy ranking and an artificial neural network in Hassi Messaoud oil field, Algeria[J]. Journal of Petroleum Science and Engineering, 2014, 115: 78-89.
[6] ITURRAR N-VIVEROS U, PARRA J O. Artificial Neural Networks applied to estimate permeability, porosity and intrinsic attenuation using seismic attributes and well-log data[J]. Journal of Applied Geophysics, 2014, 107: 45-54.
[7] WANG G, CARR T R, JU Y, et al.Identifying organic-rich Marcellus Shale lithofacies by support vector machine classifier in the Appalachian basin[J]. Computers & Geosciences, 2014, 64: 52-60.
[8] SILVA A A, NETO I A L, MISS GIA 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.
[9] SINGH U K.Fuzzy inference system for identification of geological stratigraphy off Prydz Bay, East Antarctica[J]. Journal of Applied Geophysics, 2011, 75(4): 687-698.
[10] SILVERSIDES K, MELKUMYAN A, WYMAN D, et al.Automated recognition of stratigraphic marker shales from geophysical logs in iron ore deposits[J]. Computers & Geosciences, 2015, 77: 118-125.
[11] ROLON L, MOHAGHEGH S D, AMERI S, et al.Using artificial neural networks to generate synthetic well logs[J]. Journal of Natural Gas Science and Engineering, 2009, 1(4/5): 118-133.
[12] ALIZADEH B, NAJJARI S, KADKHODAIE-ILKHCHI A.Artificial neural network modeling and cluster analysis for organic facies and burial history estimation using well log data: A case study of the South Pars Gas Field, Persian Gulf, Iran[J]. Computers & Geosciences, 2012, 45: 261-269.
[13] MO X, ZHANG Q, LI X.Well logging curve reconstruction based on Genetic Neural Networks[C]//Proceedings of the 12th International Conference on Fuzzy Systems and Knowledge Discovery (FSKD). Zhangjiajie, China: IEEE, 2015.
[14] LONG W, CHAI D, AMINZADEH F.Pseudo density log generation using artificial neural network[R]. SPE 180439-MS, 2016.
[15] SALEHI M M, RAHMATI M, KARIMNEZHAD M, et al.Estimation of the non records logs from existing logs using artificial neural networks[J]. Egyptian Journal of Petroleum, 2016, 26(4): 957-968.
[16] ADAMOWSKI J, CHAN H F.A wavelet neural network conjunction model for groundwater level forecasting[J]. Journal of Hydrology, 2011, 407(1/2/3/4): 28-40.
[17] SAHOO S, RUSSO T, ELLIOTT J, et al.Machine learning algorithms for modeling groundwater level changes in agricultural regions of the US[J]. Water Resources Research, 2017, 53(5): 3878-3895.
[18] SCHUSTER M, PALIWAL K K.Bidirectional recurrent neural networks[J]. IEEE Transactions on Signal Processing, 1997, 45(11): 2673-2681.
[19] HOCHREITER S, SCHMIDHUBER J.Long short-term memory[J]. Neural Computation, 1997, 9(8): 1735-1780.
[20] MIKOLOV T, KARAFI T 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.
[21] 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.
[22] GRAVES A, JAITLY N.Towards end-to-end speech recognition with recurrent neural networks[C]//Proceedings of the 23rd International Conference on Machine Learning. Washington D.C.: IEEE Computer Society Press, 2014.
[23] SILVA I N D, CAGNON J Â, SAGGIORO N J. Recurrent neural network based approach for solving groundwater hydrology problems[EB/OL]. (2013-01-16)[2018-05-20]. https://www.intechopen.com/books/artificial-neural-networks-architectures-and-applications/recurrent-neural-network-based-approach-for-solving-groundwater- hydrology-problems.
[24] ZHANG J, ZHU Y, ZHANG X, et al.Developing a Long Short-Term Memory (LSTM) based model for predicting water table depth in agricultural areas[J]. Journal of Hydrology, 2018, 561: 918-929.
[25] CYBENKO G.Approximation by superpositions of a sigmoidal function[J]. Mathematics of Control, Signals and Systems, 1989, 2(4): 303-314.
[26] HORNIK K.Approximation capabilities of multilayer feedforward networks[J]. Neural Networks, 1991, 4(2): 251-257.
[27] LI F F, KARPATHY A, JOHNSON J.Convolutional neural networks for visual recognition[R]. Palo Alto: Stanford University, 2015.
[28] KARPATHY A. The unreasonable effectiveness of recurrent neural networks[EB/OL]. (2015-05-21)[2018-05-26]. http://karpathy.github.io/2015/05/21/rnn-effectiveness/.
[29] OLAH C. Understanding LSTM networks[EB/OL]. (2015-08-27) [2018-05-26]. http://colah.github.io/posts/2015-08-Understanding-LSTMs/.
[30] BENGIO Y, SIMARD P, FRASCONI P.Learning long-term dependencies with gradient descent is difficult[J]. IEEE Transactions on Neural Networks, 1994, 5(2): 157-166.
[31] BREYER J A.The Eagle Ford Shale: A renaissance in U.S. oil production[M]. Tulsa: American Association of Petroleum Geologists, 2016.
[32] PARSHALL J.Drilling data provide solution to horizontal well log costs[J]. Journal of Petroleum Technology, 2015, 67(8): 35-38.
