- Diamond and Carbon-based Materials Research
- Infrastructure Maintenance and Monitoring
- Semiconductor materials and devices
- Metal and Thin Film Mechanics
- Ion-surface interactions and analysis
- Topic Modeling
- Industrial Vision Systems and Defect Detection
- Advanced Graph Neural Networks
- Data Quality and Management
- Hydraulic and Pneumatic Systems
- Fault Detection and Control Systems
- Currency Recognition and Detection
Southwest Jiaotong University
2005-2024
Due to the harsh working environment of storage stacking machinery, fault information important components is significantly complex, which leads problem low classification accuracy and high computational complexity existing deep learning-based diagnosis methods. To alleviate problem, this paper presents a novel architecture named attention-based adaptive multimodal feature fusion networks for intelligent aimed at improving diagnostic precision robustness network learning broader...
The manufacturing industry is currently experiencing a wave of digitalization and intelligent transformation. With continuous upgrades production equipment systems, the complexity diversity line faults are continuously increasing. Moreover, there weak correlation among fault-related information low utilization rate fault knowledge. In light problems, an automatic knowledge graph construction framework proposed for maintenance. employs both BERT-based models template-based methods to extract...
To address the issue of unplanned downtime caused by concrete pump piston failures and high maintenance costs extended resulting from preventative maintenance, a multi-model fusion residual life prediction method based on Stacking ensemble learning is proposed. As sample data small-batch, high-dimensional, non-linear heterogeneous data, results single model are prone to overfitting. Therefore, this study uses three regression algorithms, SVR, XGBoost, BP neural network, establish models,...