ABSTRACT
In order to make better use of the correlation between few lithological features and reduce the model complexity of convolutional neural network, this paper proposes a lithology recognition method based on a multi-scale residual one-dimensional convolutional neural network. Firstly, according to the logging data, acoustic, density, gamma ray, deep lateral resistivity, shallow lateral resistivity, photoelectric absorption cross-sectional index, p-wave velocity and shear wave velocity are selected as lithological characteristics. Due to the large difference between features and the existence of abnormal data, the Laida criterion, least squares moving average filtering and z-score standardization are used for preprocessing. Then, borrowing from the multi-scale idea of inception structure in GoogLeNet and the residual idea of ResNet, a multi-scale residual structure (MsR) is constructed, and further a multi-scale residual one-dimensional convolutional neural network (MsRNet) is constructed with MsR. Finally, lithology identification is performed by MsRNet. Through experiments with a block in Henan Oilfield, it is proved that this method has a higher lithology recognition rate than the lithology recognition methods including the k-nearest neighbor model, the product-based neural network and direct one-dimensional convolutional neural network.
- Honggen, T., Rihui, C., and Xiaoqing, Z. 2011. Well logging response to the volcaniclastic rocks of Hailar basin and application. Chinese Journal of Geophysics. 54, 2 (Feb. 2011) 534--544.Google Scholar
- YuKun, T., Hui, Z., and Sanyi, Y. 2013. Lithologic discrimination method based on Markov random-field. Chinese Journal of Geophysics. 56, 4 (April. 2013) 1360--1368.Google Scholar
- Ziyun, L., and Xianggong, W. 1989. Determination of lithology through Probability statistics. Journal of Jianghan Petroleum Institute. 11, 2 (Jun. 1989) 35--40.Google Scholar
- Zhifeng, X., and Jifeng, Y. 2008. The Application of Cluster and Discriminant Analyses in Logging Lithology Recognition. Journal of Shandong University of Science and Technology. 27, 5 (May. 2008) 10--13.Google Scholar
- Siyuan, C., and Chunsheng, L. 2002. The Application of BP Neural Network in Reservoir Predition. Progress in Geophysics. 17, 1 (Mar. 2002) 84--90.Google Scholar
- Youxiang, D., Gentian, L., and Qifeng, S. 2016. Research on convolutional neural network for resvoir parameter prediction. Journal on Communications. 37, Z1 (Oct. 2016) 1--9.Google Scholar
- Liping, Z., Hongqi, L., and Zhongguo, Y. 2018. Intelligent Logging Lithological Interpretation With Convolution Neural Networks. Petrophysics. 59, 6 (Dec. 2018) 799--810.Google Scholar
- Gang, C., Mian, C., and Guobin, H. 2020. A new method of lithology classification based on convolutional neural network algorithm by utilizing drilling string vibration data. Energies. 13, 4 (2020).Google Scholar
- Yadigar, I., and Lyudmila, S. 2019. Lithological facies classification using deep convolutional neural network. Journal of Petroleum Science and Engineering. 174 (Mar. 2019) 216--228.Google Scholar
- Christian, S., and Wei, L. 2015. Going deeper with convolutions. In IEEE Conference on Computer Vision and Pattern Recognition (Boston, MA, United states, June 07 - 12, 2015). IEEE Computer Society, 1--9.Google Scholar
- Kaiming, H., Xiangyv, Z., and Shaoqing, R. 2016. Deep Residual Learning for Image Recognition. In 29th IEEE Conference on Computer Vision and Pattern Recognition (Las Vegas, NV, United states, June 26 - July 1, 2016). IEEE Computer Society, 770--778.Google Scholar
- Sergey, I., and Christian, S. 2015. Batch normalization: accelerating deep network training byreducing internal covariate shift. In 32nd International Conference on Machine Learning (Lille, France, July 6 - 11, 2015). International Machine Learning Society, 448--456.Google Scholar
- Nitish, S., Geoffrey, H., and Alex, K. 2014. Dropout: A simple way to prevent neural networks from overfitting. Journal of Machine Learning Research. 15 (Jun. 2014) 1929--1958.Google Scholar
- Xavier, G., and Yoshua, B. 2010. Understanding the difficulty of training deep feedforward neural networks. In 13th International Conference on Artificial Intelligence and Statistics (Sardinia, Italy, May 13 - 15, 2010). Microtome Publishing, 249--256.Google Scholar
- Diederik P, K., Jimmy Lei, B. 2015. Adam: A method for stochastic optimization. In 3rd International Conference on Learning Representations (San Diego, CA, United states, May 07- 09, 2015). International Conference on Learning Representations.Google Scholar
- Matthew, G. 1972. Kn-Nearest Neighbor Classfication. In IEEE Transactions on Information Theory. IT-18, 5 (Sep. 1972) 627--630.Google Scholar
- Yanru, Q., Han, C., and Kan, R. 2016. Product-based neural networks for user response prediction. In 16th IEEE International Conference on Data Mining (Barcelona, Catalonia, Spain, December 12 - 15, 2016). Institute of Electrical and Electronics Engineers Inc., 1149--1154.Google Scholar
Index Terms
- Lithology Identification Based on Multi-scale Residual One-dimensional Convolutional Neural Network
Recommendations
Research on reservoir lithology prediction method based on convolutional recurrent neural network
Highlights- A model possessing the local information aggregation ability of CNN and the sequence feature extraction ability of RNN is designed to mine the implicit ...
AbstractConsidering that conventional reservoir prediction methods cannot fully explore the implicit relationship between seismic attributes and reservoir lithology, a deep learning lithology prediction model combining convolutional neural ...
Graphical abstractDisplay Omitted
Application on lithology recognition with BP artificial neural network
IITA'09: Proceedings of the 3rd international conference on Intelligent information technology applicationAn Artificial Neural Network (ANN) model is established to recognize the drilled formations' lithologies while drilling. The styles of output and input of ANN are designed. The nerve cells in input layer are weight of bit (WOB), speed of rotary (SOR) ...
Scene text recognition using residual convolutional recurrent neural network
Text is a significant tool for human communication, and text recognition in scene images becomes more and more important. In this paper, we propose a residual convolutional recurrent neural network for solving the task of scene text recognition. The ...
Comments