Correo Electrónico
DNINFOA - SIA
Bibliotecas
Convocatorias
Identidad U.N.
Published
Application of Random Forest method in oil and water layer identification of logging data: a case study of the Liaohe depression
Aplicación del método de Bosques Aleatorios en la identificación de las capas de petróleo y de agua durante el registro de pozo: caso de estudio en la Depresión Liaohe
DOI:
https://doi.org/10.15446/esrj.v27n1.104741Keywords:
logging data. random forest. SMOTE. oil and water layer identification (en)información de registro de pozo, método Bosques Aleatorios, método Smote, identificación de las capas de petróleo y de agua (es)
Downloads
Accurate identification of oil and water layers is the basis of qualitative evaluation of reservoir fluid properties or industrial value and selection of testing layers of the well. The traditional oil and water layer identification is mainly based on the extensive use of the well’s logging and logging data, which is inefficient and easy to leak interpretation or misinterpretation for those reservoirs with complex geological conditions. In this paper, the random forest method of machine learning is used to select the lithology, porosity, permeability, movable fluid, oil saturation, S0, S1, S2, Tmax of rock as characteristics; smote oversampling is used to expand the sample, and the packet estimation is used to establish the oil and water layer identification model. This method is simple and easy to use, not prone to severe overfitting, and can find the potential rules in the data. The classification performance is excellent, and the accuracy rate can reach more than 89.9%, which solves the problem of low accuracy in oil-water layer identification in the past.
La identificación precisa de las capas de agua y petróleo es la base de la evaluación cualitativa de las propiedades de fluido del yacimiento o de valor industrial, y de la selección de las capas de ensayo del pozo. La identificación tradicional de las capas de petróleo y agua se basa principalmente en el uso extensivo de la información ofrecida por la adquisición de registros del pozo, la cual es ineficiente y fácil de perder información o de incurrir en malinterpretación en aquellos yacimientos con condiciones geológicas complejas. En este artículo se utilizó el método de "Bosques Aleatorios (del inglés Random Forest Method)" para seleccionar la litología, porosidad, permeabilidad, fluidos móviles, saturación de petróleo, y las características de la rocas S0, S1, S2 y Tmax. El sobremuestreo con el método Smote se usó para ampliar la muestra, y el paquete de estimación se uilizó para establecer el modelo de identificación de las capas de agua y petróleo. Este método es simple y fácil de usar, además de no ser propenso a un sobreajuste severo, y puede encontrar en la información las normas potenciales que lo rigen. La clasificación del desempeño es excelente, y el índice de exactitud puede alcanzar más del 89.9 %, lo que resuelve el problema de la baja exactitud que se presenta en la identificación de las capas de petróleo y de agua.
References
Bengio, Y., Courville, A., & Vincent, P. (2012). Representation Learning: A Review and New Perspectives. ArXiv. /abs/1206.5538. https://doi.org/10.48550/arXiv.1206.5538
Breiman, L. (2001). Random forest. Machine learning, 45, 5-32 DOI: https://doi.org/10.1023/A:1010933404324
Chawla, N. V., Bowyer, K. W., & Hall, L. O. (2011). SMOTE: synthetic minority over-sampling technique. Journal of Artificial Intelligence Research, 16(1), 321-357. DOI: https://doi.org/10.1613/jair.953
Cheng, K. (2007). A Review of the Theory and Methods of Statistical Data Preprocessing. Statistics and Information Forum, 22(6), 98-103.
Cutler, A., Cutler, D. R., & Stevens, J. R. (2004). Random Forests. Machine Learning, 45(1), 157-176. DOI: https://doi.org/10.1007/978-1-4419-9326-7_5
Džeroski, S., & Ženko, B. (2004). Is Combining Classifiers with Stacking Better than Selecting the Best One? Machine Learning, 54, 255–273. https://doi.org/10.1023/B:MACH.0000015881.36452.6e DOI: https://doi.org/10.1023/B:MACH.0000015881.36452.6e
Hang, L. (2012). Statistical learning methods. Beijing: Tsinghua University Press.
Kang, Q., & Lu, L. (2020). Application of stochastic forest algorithm in lithology classification of logging. World Geology, 39(2), 398-405.
Lai, Q., Wei, B., & Wu, Y. (2021). K-Neighbor Algorithm For Igneous Lithology Based on Random Forest. Special Oil and Gas Reservoirs, 28(6), 62-69.
Liang, J., Chen, J., & Zhang, X. (2019). Anomaly Detection Based on Durtific Coding and Convolutional Neural Network. Journal of Tsinghua University (Natural Science Edition), 59(7), 523-529.
Liu, Y., Liu, S., & Ma, Q. (2019). Application of BP neural network method in slate facies identification of Lucaogou Formation in Santanghu Basin. Lithological Reservoirs, 31(4), 101-111.
Pedregosa, F., Varoquaux, G., & Gramfort, A. (2012). Scikit-learn: Machine learning in python. Journal of Machine Learning Research, 12(10), 2825-2830.
Su, G. (2006). Application of Geochemical Gas Logging Data in Oil-Water Reservoir Identification. Logging Technology, 30(6), 551-553.
Wu, Z., Zhang, X., Zhang, C., & Wang, H. (2021). Lithology Recognition Method Based on LSTM Recurrent Neural Network. Lithological Reservoirs, 33(3),120-128.
Xing, C., Zhou, C., & He, Y. (2022). Direct inversion of pore pressure in unconventional reservoir formations by Bayesian method. Lithological Reservoirs, 34(3), 1-7.
Wang, Y., Wang, M., & Tian, S. (2021). Coal Rock Identification Based on Kalman Filter and Random Forest. Coal Technology, 40(12), 208-211.
Wang, Y., Wang, R., & Wie, K. (2021). Classification of compact reservoirs based on random forests: A case study of the eastern box 8 section of Yan'an gas field. Journal of Xi'an Shiyou University (Natural Science Edition), 36(6), 1-8.
Zhao, M., Jin, Y., & Wang, Y. (2021). Application of Stochastic Forest Algorithm in Selection Decision. Computer and Network, 47(22), 56-59.
Zhong, Y., Zhang, T., & Li, P. (2022). Study on the Classification of Stochastic Forest Fusion Model in The Classification of Pressure Well Methods. Journal of Southwest Petroleum University (Natural Science Edition), 44(1), 165-173.
Zhou, Z. (2016). Machine Learning. Beijing: Tsinghua University Press.
Zhou, X., Zhang, Z., & Zhang, C. (2017). Complex lithology recognition based on rough set-random forest algorithm. Daqing Petroleum Geology and Development, 36(6), 127-133.
How to Cite
APA
ACM
ACS
ABNT
Chicago
Harvard
IEEE
MLA
Turabian
Vancouver
Download Citation
CrossRef Cited-by
1. Tarun Jaiswal, Sujata Dash, Ganpati Panda, Sudipta Patowary, Shanchamo Yanthan. (2025). Biologically Inspired Techniques in Many Criteria Decision-Making. Learning and Analytics in Intelligent Systems. 45, p.41. https://doi.org/10.1007/978-3-031-82706-8_5.
2. Aditya Pramada Wicaksono, Achmad Choiruddin. (2024). Candidate Selection of Water Shut-Off in Oil and Gas Industry Using Random Forest. 2024 IEEE International Symposium on Consumer Technology (ISCT). , p.464. https://doi.org/10.1109/ISCT62336.2024.10791205.
Dimensions
PlumX
Article abstract page views
Downloads
License
This work is licensed under a Creative Commons Attribution 4.0 International License.
Earth Sciences Research Journal holds a Creative Commons Attribution license.
You are free to:
Share — copy and redistribute the material in any medium or format
Adapt — remix, transform, and build upon the material for any purpose, even commercially.
The licensor cannot revoke these freedoms as long as you follow the license terms.
