A5103207616- Welding Techniques and Residual Stresses
- Additive Manufacturing Materials and Processes
- Metallurgy and Material Science
- Reliability and Maintenance Optimization
- Semantic Web and Ontologies
- Quality and Safety in Healthcare
- Fault Detection and Control Systems
- Machine Fault Diagnosis Techniques
- Coal and Its By-products
- Advanced Data Processing Techniques
- Bauxite Residue and Utilization
- Industrial Vision Systems and Defect Detection
- Recycling and utilization of industrial and municipal waste in materials production
- Biomedical Text Mining and Ontologies
- Brain Tumor Detection and Classification
- Topic Modeling
Sinopec (China)
2025
Shanghai Jiao Tong University
2024
Zhejiang University of Technology
2023-2024
Abstract Efficiently classifying potential areas of remaining oil is essential for enhancing the recovery in high water cut reservoir. The distribution complex and challenging to mobilize due temporal evolution spatial variation long-term waterflood development. Currently, reservoir classification relies on manual experience unsupervised machine learning, both which have limitations. Manual constrained by human understanding, leading inaccuracies, while learning lacks adherence theory,...
Remaining useful life (RUL) prediction is of great significance to ensure the safety and reliability equipment. Graph neural network-based methods show potential improve RUL performance by extracting spatiotemporal features from sensor monitoring data. However, current construct sensor-based homogeneous graphs without considering equipment component structure data prior knowledge, which cannot characterize dependency between sensors studied accurately. To solve this problem, we propose a...
<title>Abstract</title> Weld quality inspection is essential in modern manufacturing, requiring the automatic identification, localization, and measurement of defects industrial environments. Although 2D images 3D point clouds each have their unique advantages, most current methods focus on only one these data types. This study proposes a novel system integrating cloud with using PointNet + YOLOv5. The mapped into corresponding feature maps trained separately. Training results show that...