Wei Cheng
Research Subjects
Wei Cheng has a total of 5 co-authors. Here are the top co-authors by number of shared publications:
- Fansheng Xiong
Beijing Institute of Mathematical Sciences and Applications Beijing China
1 shared publication - Jing Ba
School of Earth Sciences and Engineering, Hohai University, Nanjing, China
1 shared publication - Jose' M. Carcione
National Institute of Oceanography and Applied Geophysics — OGS 3 , Trieste, Italy
1 shared publication - Qiang Guo
China University of Mining and Technology 1 , School of Resources and Geosciences, Xuzhou, China
1 shared publication - Zhijian Fang
Hohai University 1 , School of Earth Sciences and Engineering, Nanjing, China and , Institute of Earth Sciences, Lausanne, Switzerland.
1 shared publication
Latest Publications
- 26 November 2025
- Published by Society of Exploration Geophysicists in Geophysics
- Vol. 91 (1) , WA35-WA46
Deep neural networks for estimating S-wave velocity in tight sandstone reservoirs from log data
Accurate estimation of S-wave velocity from well logs is relevant for processing seismic data from hydrocarbon reservoirs, CO2 storage sites, and geothermal systems. Standard methods rely on empirical relationships and basic rock-physics models (RPMs), such as the Gassmann equation, which are often inaccurate because they cannot account for the complex factors that influence the relationship between S-wave velocity and other properties. Machine-learning techniques, such as deep neural networks (DNNs), have been shown to be effective methods for the indirect estimation of S-wave logs and other rock properties. In this work, we develop a novel prediction method that combines RPMs (Xu-White and Biot-Rayleigh [BR] models) and DNNs based on pseudolabels, i.e., a class of algorithms that use model outputs to assign labels to unlabeled samples, which are then used as labeled samples during training. The BR model takes into account the volume ratio of inclusions and treats it as a spatially varying parameter, allowing for a better description of the pore microstructure. Data from five boreholes demonstrate the applicability of our model. A single well is used for training, and the remaining four wells are used as a blind test. Compared with DNNs without physical poroelasticity models, the method driven by models indicates a significant improvement in prediction accuracy. By incorporating pseudolabels into the BR model, the prediction performance of the network model is improved. The method proposed in this work is validated and applied in other work areas.
- 26 November 2025
- Published by Society of Exploration Geophysicists in Geophysics
- Vol. 91 (1) , WA35-WA46
Deep neural networks for estimating S-wave velocity in tight sandstone reservoirs from log data
Accurate estimation of S-wave velocity from well logs is relevant for processing seismic data from hydrocarbon reservoirs, CO2 storage sites, and geothermal systems. Standard methods rely on empirical relationships and basic rock-physics models (RPMs), such as the Gassmann equation, which are often inaccurate because they cannot account for the complex factors that influence the relationship between S-wave velocity and other properties. Machine-learning techniques, such as deep neural networks (DNNs), have been shown to be effective methods for the indirect estimation of S-wave logs and other rock properties. In this work, we develop a novel prediction method that combines RPMs (Xu-White and Biot-Rayleigh [BR] models) and DNNs based on pseudolabels, i.e., a class of algorithms that use model outputs to assign labels to unlabeled samples, which are then used as labeled samples during training. The BR model takes into account the volume ratio of inclusions and treats it as a spatially varying parameter, allowing for a better description of the pore microstructure. Data from five boreholes demonstrate the applicability of our model. A single well is used for training, and the remaining four wells are used as a blind test. Compared with DNNs without physical poroelasticity models, the method driven by models indicates a significant improvement in prediction accuracy. By incorporating pseudolabels into the BR model, the prediction performance of the network model is improved. The method proposed in this work is validated and applied in other work areas.