A5103248433- Enhanced Oil Recovery Techniques
- Hydraulic Fracturing and Reservoir Analysis
- Hydrocarbon exploration and reservoir analysis
- Vibration and Dynamic Analysis
- Reservoir Engineering and Simulation Methods
- Domain Adaptation and Few-Shot Learning
- CO2 Sequestration and Geologic Interactions
- Simulation and Modeling Applications
- Aerodynamics and Fluid Dynamics Research
- Human Pose and Action Recognition
- Multimodal Machine Learning Applications
- Advanced Topology and Set Theory
- Fluid Dynamics and Vibration Analysis
- Groundwater flow and contamination studies
- Industrial Technology and Control Systems
- Advanced Neural Network Applications
- Petroleum Processing and Analysis
- CCD and CMOS Imaging Sensors
- Manufacturing Process and Optimization
- Advanced Measurement and Detection Methods
- NMR spectroscopy and applications
- Advanced Numerical Analysis Techniques
- Industrial Vision Systems and Defect Detection
- Visual Attention and Saliency Detection
- Fire Detection and Safety Systems
Institute of Automation
2022-2024
Chinese Academy of Sciences
2021-2024
University of Chinese Academy of Sciences
2022-2024
Hanyang University
2024
Petro-Canada
2020-2022
Beijing Academy of Artificial Intelligence
2022
Hohai University
2021
Southwest Jiaotong University
2018
Guangzhou University
2017
South China University of Technology
2016
Inspired by the water-collecting mechanism of Stenocara beetle's back structure, we prepared a superhydrophilic bumps-superhydrophobic/superoleophilic stainless steel mesh (SBS-SSM) filter via facile and environmentally friendly method. Specifically, hydrophilic silica microparticles are assembled on as-cleaned surface, followed further spin-coating with fluoropolymer/SiO2 nanoparticle solution. On special surface SBS-SSM, attributed to steep energy gradient, bumps (hydrophilic...
Densities and solubilities for binary systems of carbon dioxide + water Weyburn-formation brine were measured at a temperature 59 °C pressures up to 29 MPa. Density values obtained both the saturated unsaturated aqueous solutions different pressures. A correlation in literature was used predict solubility phase, results compared with data. The density correlated as function concentration pressure. simple method determining recommended.
For both CO2 enhanced oil recovery and sequestration in reservoirs, the diffusion of injected into oil-saturated porous media is great significance project design, risk assessments, economic evaluation, performance forecast. However, measurement coefficient liquid-saturated under reservoir conditions has not been well-established due to complications caused by density-induced natural convection CO2−oil systems. This paper presented a new method for measuring effective conditions. A core plug...
Inspired by the masked language modeling (MLM) in natural processing tasks, image (MIM) has been recognized as a strong self-supervised pre-training method computer vision. However, high random mask ratio of MIM results two serious problems: 1) inadequate data utilization images within each iteration brings prolonged pre-training, and 2) inconsistency predictions unreliable generations, i.e., prediction identical patch may be inconsistent different rounds, leading to divergent semantics...
Diffusive tortuosity factor is one of the key parameters in modeling solute diffusion liquid-saturated porous media. However, determination diffusive has to involve a process media, which was usually found complicated laboratories. The incorrect use may cause significant errors certain circumstances. This paper presents method evaluate factors for consolidated i.e., sandstones, are typical media commonly encountered contaminant transport underground water and gas migration reservoirs....
Transformer has demonstrated exceptional performance on a variety of vision tasks. However, its high computational complexity can become problematic. In this paper, we conduct systematic analysis the each component in transformers, and identify an easily overlooked detail: that Feed-Forward Network (FFN) is primary bottleneck, even more so than Multi-Head Self-Attention (MHSA) mechanism. Inspired by this, further propose lightweight FFN module, named SparseFFN, reduce dense computations both...
This paper focuses on the development of natural convection mass transfer induced by gas diffusion in a liquid-saturated vertical porous column. The column is confined with two impermeable end faces and an isosolutal side surface. Mass studied numerically solving conservation equations mass, momentum, diffusing species. To describe transient behavior convection, integrated Sherwood numbers, i.e., Sht Shi, are defined basis parallel pure case. evolution process systematically for aspect...
A new method for determining the gas effective diffusion coefficient in brine-saturated porous rocks was developed on basis of radial model. The experiment conducted a high-pressure cell containing column middle and annulus between cell. two end faces were sealed diffused into only through direction. small-pressure decay technique employed to measure amount as function time. mathematical model derived determine from measured pressure decay. To test presented method, eight measurements made...
Summary Use of oil-in-water (O/W) emulsion has shown its potential for conformance control in heterogeneous porous media, yet it is essential to understand how improve the performance 2D model with lean zones before applied fields. In this paper, an O/W emulsion-based method improved through newly designed flow tests and optimized modeling study. A was a high water mobility zone (HWMZ) low (LWMZ) separated by horizontal injection well mimic real oil sands (top- or bottomwater) application...
Visual tasks vary a lot in their output formats and concerned contents, therefore it is hard to process them with an identical structure. One main obstacle lies the high-dimensional outputs object-level visual tasks. In this paper, we propose object-centric vision framework, Obj2Seq. Obj2Seq takes objects as basic units, regards most sequence generation problems of objects. Therefore, these can be decoupled into two steps. First recognize given categories, then generate for each The...
Previous unsupervised domain adaptation methods did not handle the cross-domain problem from perspective of frequency for computer vision. The images or feature maps different domains can be decomposed into low-frequency component and high-frequency component. This paper pro-poses assumption that information is more domain-invariant while con-tains domain-related information. Hence, we introduce an approach, named module (LFM), to extract representations. LFM constructed with digital...
Abstract This paper reports the laboratory measurements of gas permeability for tight reservoir core samples with various upstream, downstream, and average pressures. Measurements are conducted backpressures from atmospheric pressure up to 9.3 MPa. It is found that exerting a backpressure at outlet sample can effectively reduce slip effect improve measurements. When reaches certain level, defined as minimum backpressure, be eliminated so non-slip obtained without using Klinkenberg...
Shape reconstruction of aerospace plate structure is an important issue for safe operation vehicles. One way to achieve such by constructing smart fiber Bragg grating (FBG) with discrete distributed FBG sensor arrays using algorithms in which error analysis algorithm a key link. Considering that traditional methods can only deal static data, new dynamic data method are proposed based on LMS shape structure. Firstly, and orthogonal curved network introduced. Then, model analysis. Thirdly, the...
Determination of Gas Permeability Tight Reservoir Cores without Using Klinkenberg Correlation Shuliang Li; Li University Regina Search for other works by this author on: This Site Google Scholar Mingzhe Dong; Dong Liming Dai; Dai Zhaowen Xuefeng Pan Core Laboratories Paper presented at the SPE Asia Pacific Oil and Conference Exhibition, Perth, Australia, October 2004. Number: SPE-88472-MS https://doi.org/10.2118/88472-MS Published: 18 2004 Cite View Citation Add to Manager Share Icon Twitter...