A5100720998- Hydrocarbon exploration and reservoir analysis
- Enhanced Oil Recovery Techniques
- Fault Detection and Control Systems
- Neural Networks and Applications
- NMR spectroscopy and applications
- Robotic Mechanisms and Dynamics
- Machine Fault Diagnosis Techniques
- Robot Manipulation and Learning
- Control Systems and Identification
- Petroleum Processing and Analysis
- Blind Source Separation Techniques
- Visual Attention and Saliency Detection
- Structural Health Monitoring Techniques
- Advanced Sensor and Control Systems
- Industrial Vision Systems and Defect Detection
- Industrial Technology and Control Systems
- Geoscience and Mining Technology
- Advanced Manufacturing and Logistics Optimization
- Advanced Neural Network Applications
- Magnetic confinement fusion research
- Hydraulic Fracturing and Reservoir Analysis
- Reinforcement Learning in Robotics
- Belt Conveyor Systems Engineering
- Robotic Path Planning Algorithms
- Advanced Algorithms and Applications
Chinese Academy of Sciences
2024
Institute of Information Engineering
2024
Shandong University of Technology
2022
Sinopec (China)
2013-2021
China University of Petroleum, Beijing
2017
Henan Normal University
2013
Southwestern Institute of Physics
2012
Anhui Agricultural University
2007
Fabric defect detection is a challenging task in the fabric industry because of complex shapes and large variety defects.Many methods have been proposed to solve this problem, but their speed accuracy were very low.As classic deep learning method end-to-end target algorithm, YOLOv4 has evolved rapidly applied many industries, showing good performance.This paper proposes an improved algorithm with higher for detection, which new SPP structure that uses SoftPool instead MaxPool adopted.The...
Open-Set Recognition (OSR) aims to accurately identify known classes while effectively rejecting unknown guarantee reliability. Most existing OSR methods focus on learning in the spatial domain, where subtle texture and global structure are potentially intertwined. Empirical studies have shown that DNNs trained original domain inclined over-perceive texture. The biased semantic perception could lead catastrophic over-confidence when predicting both classes. To this end, we propose an...
Abstract The formation mechanism and utilization conditions of the remaining oil in high water cut period play significant roles improved tapping potential enhanced recovery. classification is a difficult point, meanwhile burning issue. However, current method mainly through manual to determine boundaries classification, time-consuming has great subjectivity. Machine learning data mining methods recent years have been widely used field petroleum engineering, such as prediction recovery...
For reservoirs in high water cut exploitation period, profile control and plugging is one of the important ways to improve oil recovery. Cores with different permeability were flooded analyze displacement results mechanisms for grain size polymer microspheres. Fluid distribution cores was measured by NMR spectroscopy after flooding, microspheres subsequent flooding. The range pore sizes from which swept out also calculated. showed that can effectively sweep remaining cores. suitability...
In addition to the real time plasma magnetic and kinetic control, process control is also very important for operation of a Tokamak. Process aims at normal plant systems including information sharing, schema Tokamak operation, alarm handling. A new tokamak named “HL-2M” will be built during next four years many conceptual designs device might validated on HL-2A, which currently running SWIP Chengdu, China. For those purposes, we select EPICS as basis system HL-2A. Some user interface, data...
Abstract For reservoirs containing oil with a high total acid number, alkali-cosolvent-polymer (ACP) flood can potentially increase the recovery by its saponification effects. The enhanced performance of ACP has been studied at core and reservoir scale in detail, however, effect on residual saturation swept area still lacks enough research. Medical computed tomography (Medical-CT) scan micro (Micro-CT) are used combination to visualize micro-scale flow reveal mechanisms reduction during...
The ability to swiftly and precisely reach any user-specified target location is necessary for a robotic arm that can be used in real-world scenarios. To date, many evolutionary optimization algorithms have been design controllers arms. However, when designing multiple targets, most existing methods need evolve the control strategy from scratch each target, rather than trying reuse experience. Therefore, computational resources are repeatedly meaninglessly consumed. this end, paper proposes...