Well Log Generation via Ensemble Long Short-Term Memory (EnLSTM) Network

被引:38
作者
Chen, Yuntian [1 ]
Zhang, Dongxiao [2 ]
机构
[1] Frontier Res Ctr, Intelligent Energy Lab, Peng Cheng Lab, Shenzhen, Peoples R China
[2] Southern Univ Sci & Technol, Sch Environm Sci & Engn, Shenzhen, Peoples R China
基金
中国国家自然科学基金;
关键词
few-shot learning; sequential data; well log; logging-while-drilling; LSTM; ARTIFICIAL NEURAL-NETWORKS; RESERVOIR CHARACTERIZATION; POROSITY; PROJECT; ROCKS;
D O I
10.1029/2020GL087685
中图分类号
P [天文学、地球科学];
学科分类号
07 ;
摘要
In this study, we propose an ensemble long short-term memory (EnLSTM) network, which can be trained on a small data set and process sequential data. The EnLSTM is built by combining the ensemble neural network and the cascaded LSTM network to leverage their complementary strengths. Two perturbation methods are applied to resolve the issues of overconvergence and disturbance compensation. The EnLSTM is compared with commonly used models on a published data set and proven to be the state-of-the-art model in generating well logs. In the case study, 12 well logs that cannot be measured while drilling are generated based on the logs available in the drilling process. The EnLSTM is capable of reducing cost and saving time in practice. Plain Language Summary A novel neural network, called EnLSTM, is proposed by combining the ensemble neural network, which has good performance on small-data problems, and the cascaded long short-term memory network, which is effective at processing sequential data. The EnLSTM's capability of processing sequential data based on a small data set is especially suitable for generating synthetic well logs. In addition, two perturbation methods are used to ensure that the EnLSTM can be fully trained in practice. In the experiments, the EnLSTM achieved the current best results on a published well log data set, and its application value is verified in a case study. Key Points We proposed an ensemble long short-term memory (EnLSTM) network to process sequential data based on a small dataset The EnLSTM solved a well log generation problem with higher prediction accuracy than the previously best model on a published dataset The EnLSTM accurately generated 12 hard-to-measure well logs based on LWD logs, resulting in a reduction of cost and time in practice
引用
收藏
页数:9
相关论文
共 38 条
[1]  
Akinnikawe O., 2018, UNC RES TECHN C HOUS
[2]  
Alexeyev A., 2017, Well Log Based Geomechanical and Petrophysical Analysis of the Bakken Formation
[3]  
Aron J., 1994, SPWLA 35 ANN LOGG S, P1
[4]  
Bertsekas D. P., 1998, NONLINEAR PROGRAMMIN
[5]   Surrogate model based iterative ensemble smoother for subsurface flow data assimilation [J].
Chang, Haibin ;
Liao, Qinzhuo ;
Zhang, Dongxiao .
ADVANCES IN WATER RESOURCES, 2017, 100 :96-108
[6]   Levenberg-Marquardt forms of the iterative ensemble smoother for efficient history matching and uncertainty quantification [J].
Chen, Yan ;
Oliver, Dean S. .
COMPUTATIONAL GEOSCIENCES, 2013, 17 (04) :689-703
[7]   Physics-Constrained Deep Learning of Geomechanical Logs [J].
Chen, Yuntian ;
Zhang, Dongxiao .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2020, 58 (08) :5932-5943
[8]   Ensemble Neural Networks (ENN): A gradient-free stochastic method [J].
Chen, Yuntian ;
Chang, Haibin ;
Meng, Jin ;
Zhang, Dongxiao .
NEURAL NETWORKS, 2019, 110 :170-185
[9]   Leave one out error, stability, and generalization of voting combinations of classifiers [J].
Evgeniou, T ;
Pontil, M .
MACHINE LEARNING, 2004, 55 (01) :71-97
[10]   Reservoir characterization of a CO2 storage aquifer: The Upper Triassic Stuttgart Formation in the Northeast German Basin [J].
Foerster, A. ;
Schoener, R. ;
Foerster, H. -J. ;
Norden, B. ;
Blaschke, A. -W. ;
Luckert, J. ;
Beutler, G. ;
Gaupp, R. ;
Rhede, D. .
MARINE AND PETROLEUM GEOLOGY, 2010, 27 (10) :2156-2172