Wenhao Fu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Wenhao Fu has a total of 24 co-authors. Here are the top co-authors by number of shared publications:
- Kai Zhang
Civil Engineering School, Qingdao University of Technology, Qingdao, China
4 shared publications - Piyang Liu
School of Civil Engineering, Qingdao University of Technology, Qingdao, China
3 shared publications - Hai Sun
School of Petroleum Engineering, China University of Petroleum (East China), Qingdao 266580, China
2 shared publications - Jinding Zhang
School of Petroleum Engineering, China University of Petroleum (East China), Qingdao, China
2 shared publications Jun Yao
School of Petroleum Engineering, China University of Petroleum (East China), Qingdao, China
2 shared publicationsLiming Zhang
2 shared publications
Latest Publications
- 28 January 2026
- Published by Elsevier in Journal of Hydrology
- Vol. 668
Parameterization of complex geological models with PCA‑guided adversarial diffusion for ensemble data assimilation
- 28 January 2026
- Published by Elsevier in Journal of Hydrology
- Vol. 668
Parameterization of complex geological models with PCA‑guided adversarial diffusion for ensemble data assimilation
- 2 December 2025
- Published by Society of Petroleum Engineers (SPE) in SPE Journal
- Vol. 31 (02) , 697-711
Parameter Inversion in Latent Space for Large-Scale Reservoirs via Deep Generative Adversarial Networks
Summary: Parameter inversion plays an important role in many science and engineering fields, typically involving the process of determining the internal parameters of a system based on actual observational data. In recent years, deep learning methods have been widely applied to inversion problems, significantly improving computational efficiency. The end-to-end inversion method avoids the iterative adjustment of parameters and greatly improves the computing efficiency. However, using existing end-to-end methods to perform parameter inversion for large-scale 3D reservoir models is still a challenge. Therefore, to achieve end-to-end inversion for large-scale 3D reservoirs, we propose a new inversion framework. Specifically, the data space inversion (DSI) method is used to predict posterior dynamic data, reducing model uncertainty. Dimensionality reduction (DR) methods transform high-dimensional model parameters into low-dimensional latent variables. By training the image-to-image generative adversarial network (pix2pixGAN), the framework seeks to establish a mapping relationship between dynamic data and low-dimensional latent variables. We tested this method (DSI-DR-pix2pixGAN) on 2D and 3D groundwater flow models and compared the proposed approach with the DSI-pix2pixGAN, DSI-DR-U-Net, and ensemble smoother (ES) with multiple data assimilation (ES-MDA). The results indicate that the proposed method performs well in the large-scale 3D model, significantly reducing computational costs while achieving effective outcomes.
- 2 December 2025
- Published by Society of Petroleum Engineers (SPE) in SPE Journal
- Vol. 31 (02) , 697-711
Parameter Inversion in Latent Space for Large-Scale Reservoirs via Deep Generative Adversarial Networks
Summary: Parameter inversion plays an important role in many science and engineering fields, typically involving the process of determining the internal parameters of a system based on actual observational data. In recent years, deep learning methods have been widely applied to inversion problems, significantly improving computational efficiency. The end-to-end inversion method avoids the iterative adjustment of parameters and greatly improves the computing efficiency. However, using existing end-to-end methods to perform parameter inversion for large-scale 3D reservoir models is still a challenge. Therefore, to achieve end-to-end inversion for large-scale 3D reservoirs, we propose a new inversion framework. Specifically, the data space inversion (DSI) method is used to predict posterior dynamic data, reducing model uncertainty. Dimensionality reduction (DR) methods transform high-dimensional model parameters into low-dimensional latent variables. By training the image-to-image generative adversarial network (pix2pixGAN), the framework seeks to establish a mapping relationship between dynamic data and low-dimensional latent variables. We tested this method (DSI-DR-pix2pixGAN) on 2D and 3D groundwater flow models and compared the proposed approach with the DSI-pix2pixGAN, DSI-DR-U-Net, and ensemble smoother (ES) with multiple data assimilation (ES-MDA). The results indicate that the proposed method performs well in the large-scale 3D model, significantly reducing computational costs while achieving effective outcomes.
- 18 April 2025
- Published by Elsevier in Engineering Applications of Artificial Intelligence
- Vol. 153
Use deep transfer learning for efficient time-series updating of subsurface flow surrogate model
- 18 April 2025
- Published by Elsevier in Engineering Applications of Artificial Intelligence
- Vol. 153
Use deep transfer learning for efficient time-series updating of subsurface flow surrogate model
- 2 June 2023
- Published by Society of Petroleum Engineers (SPE) in SPE Journal
- Vol. 28 (05) , 2223-2239
Inversion Framework of Reservoir Parameters Based on Deep Autoregressive Surrogate and Continual Learning Strategy
Summary: History matching is a crucial process that enables the calibration of uncertain parameters of the numerical model to obtain an acceptable match between simulated and observed historical data. However, the implementation of the history-matching algorithm is usually based on iteration, which is a computationally expensive process due to the numerous runs of the simulation. To address this challenge, we propose a surrogate model for simulation based on an autoregressive model combined with a convolutional gated recurrent unit (ConvGRU). The proposed ConvGRU-based autoregressive neural network (ConvGRU-AR-Net) can accurately predict state maps (such as saturation maps) based on spatial and vector data (such as permeability and relative permeability, respectively) in an end-to-end fashion. Furthermore, history matching must be performed multiple times throughout the production cycle of the reservoir to fit the most recent production observations, making continual learning crucial. To enable the surrogate model to quickly learn recent data by transferring experience from previous tasks, an ensemble-based continual learning strategy is used. Together with the proposed neural network–based surrogate model, the randomized maximum likelihood (RML) is used to calibrate uncertain parameters. The proposed method is evaluated using 2D and 3D reservoir models. For both cases, the surrogate inversion framework successfully achieves a reasonable posterior distribution of reservoir parameters and provides a reliable assessment of the reservoir’s behaviors.
- 2 June 2023
- Published by Society of Petroleum Engineers (SPE) in SPE Journal
- Vol. 28 (05) , 2223-2239
Inversion Framework of Reservoir Parameters Based on Deep Autoregressive Surrogate and Continual Learning Strategy
Summary: History matching is a crucial process that enables the calibration of uncertain parameters of the numerical model to obtain an acceptable match between simulated and observed historical data. However, the implementation of the history-matching algorithm is usually based on iteration, which is a computationally expensive process due to the numerous runs of the simulation. To address this challenge, we propose a surrogate model for simulation based on an autoregressive model combined with a convolutional gated recurrent unit (ConvGRU). The proposed ConvGRU-based autoregressive neural network (ConvGRU-AR-Net) can accurately predict state maps (such as saturation maps) based on spatial and vector data (such as permeability and relative permeability, respectively) in an end-to-end fashion. Furthermore, history matching must be performed multiple times throughout the production cycle of the reservoir to fit the most recent production observations, making continual learning crucial. To enable the surrogate model to quickly learn recent data by transferring experience from previous tasks, an ensemble-based continual learning strategy is used. Together with the proposed neural network–based surrogate model, the randomized maximum likelihood (RML) is used to calibrate uncertain parameters. The proposed method is evaluated using 2D and 3D reservoir models. For both cases, the surrogate inversion framework successfully achieves a reasonable posterior distribution of reservoir parameters and provides a reliable assessment of the reservoir’s behaviors.
- 16 March 2023
- Published by American Geophysical Union (AGU) in Water Resources Research
- Vol. 59 (3)
Deep Conditional Generative Adversarial Network Combined With Data‐Space Inversion for Estimation of High‐Dimensional Uncertain Geological Parameters
Inverse modeling can provide a reliable geological model for subsurface flow numerical simulation, which is a challenging issue that requires calibration of the uncertain parameters of the geological model to establish an acceptable match between simulation data and observation data. The general inverse modeling method needs to iteratively adjust the uncertain parameters, which is a difficult and time‐consuming high‐dimensional sampling problem. To address this problem, we propose a deep‐learning‐based inverse modeling method called pix2pixGAN‐DSI. In this method, the deep‐learning‐based image‐to‐image generative adversarial network (pix2pixGAN) is constructed to directly predict the posterior parameter fields from the posterior dynamic responses obtained by the data‐space inversion (DSI) method. This inverse modeling method does not need to iteratively adjust the uncertain parameters, which improves computational efficiency. The effectiveness of the proposed method is verified through a Gaussian model case and two non‐Gaussian channelized model cases. Through the analysis of posterior realizations, matching and forecast of production data, and uncertainty quantification, the results show that the proposed method can obtain reasonable estimates without iteration and parameterization.
- 16 March 2023
- Published by American Geophysical Union (AGU) in Water Resources Research
- Vol. 59 (3)
Deep Conditional Generative Adversarial Network Combined With Data‐Space Inversion for Estimation of High‐Dimensional Uncertain Geological Parameters
Inverse modeling can provide a reliable geological model for subsurface flow numerical simulation, which is a challenging issue that requires calibration of the uncertain parameters of the geological model to establish an acceptable match between simulation data and observation data. The general inverse modeling method needs to iteratively adjust the uncertain parameters, which is a difficult and time‐consuming high‐dimensional sampling problem. To address this problem, we propose a deep‐learning‐based inverse modeling method called pix2pixGAN‐DSI. In this method, the deep‐learning‐based image‐to‐image generative adversarial network (pix2pixGAN) is constructed to directly predict the posterior parameter fields from the posterior dynamic responses obtained by the data‐space inversion (DSI) method. This inverse modeling method does not need to iteratively adjust the uncertain parameters, which improves computational efficiency. The effectiveness of the proposed method is verified through a Gaussian model case and two non‐Gaussian channelized model cases. Through the analysis of posterior realizations, matching and forecast of production data, and uncertainty quantification, the results show that the proposed method can obtain reasonable estimates without iteration and parameterization.