A5065794169- Railway Systems and Energy Efficiency
- Advanced Battery Technologies Research
- Image Enhancement Techniques
- Traffic Prediction and Management Techniques
- Silicon Carbide Semiconductor Technologies
- Multilevel Inverters and Converters
- Semantic Web and Ontologies
- Reliability and Maintenance Optimization
- Emotion and Mood Recognition
- Railway Engineering and Dynamics
- Power Systems and Technologies
- Rough Sets and Fuzzy Logic
- Data Quality and Management
- Advanced Image Fusion Techniques
- Data Management and Algorithms
- EEG and Brain-Computer Interfaces
- Transportation Planning and Optimization
- Occupational Health and Safety Research
- Electric Vehicles and Infrastructure
- Video Surveillance and Tracking Methods
- Engineering Diagnostics and Reliability
- Advanced Battery Materials and Technologies
- Fire Detection and Safety Systems
- Topic Modeling
- Machine Fault Diagnosis Techniques
Central South University
2020-2023
Ministry of Transport
2023
Captured images of outdoor scenes usually exhibit low visibility in cases severe haze, which interferes with optical imaging and degrades image quality. Most the existing methods solve single-image dehazing problem by applying supervised training on paired images; however, practice, pairing real-world is not viable. Additionally, processing speed individual models important practical applications. In this study, a novel unsupervised single network (USID-Net) based disentangled...
The intelligent diagnosis model based on deep learning method can effectively and accurately diagnose the health state of bearings widely used in various mechanical equipment. However, engineering practice, it is difficult to obtain sufficient labeled fault data, which would reduce performance. In order solve above problem, this article proposes a self-attention metric multiscale feature fusion classify bearing with few shots. proposed few-shot mainly contains extraction module module....
Captured outdoor scene images are easily affected by haze. Most image dehazing methods have limited generalization capabilities for real-world hazy owing to the complexities of environments and domain gaps in training datasets. This article proposes a semi-supervised single-image network based on disentangled meta-knowledge. The symmetric heterogeneous design is conducive separation content mask features these used as meta-knowledge guide feature fusion network. Moreover, functions...
To achieve precise prediction of the air pressure changes in critical components brake system, including pipes (BP), auxiliary reservoirs (AR), and cylinders (BC), this paper proposed a neural network-based model developed from application release experimental data with different applications (50, 70, 100, 150 kPa) conducted on heavy haul train (HHT) test platform. The physical structures system working mechanisms network are introduced, scheme is presented by analysing variation within...
Abstract Trip destination prediction plays an important role in exploring urban travel patterns. Accurate can improve the efficiency of traffic management and quality location‐based services. Here, a deep learning structure that contains three components: information extraction, classification mechanism, output module is proposed. Three types (the partial trajectory on‐going trips, historical trajectories, related external information) are extracted first component. Then, mechanism chooses...
DC-link capacitor is one of the significant parts traction converters. Due to impact electrical stress, high temperature and humidity, its capacitance degrades faster than expected. Accurate estimation critical condition monitoring capacitors, which also basis online life prediction. Currently, there a big challenge for in railway applications. The noise fluctuation voltage sensor may be nearly equal that ripple voltage, leading considerable errors existing methods. Therefore, method based...
A dc-link capacitor is an important device and the vulnerable electronic part of train converter. The capacitance decline accelerated because influence voltage, current, charge–discharge frequency, temperature, other factors. This will lead to circuit performance degradation even failures. Therefore, estimation capacitors necessary great significance safe operation train. However, voltage sensors used in converter have large noise low sampling which makes it difficult obtain accurate...
Co-phase power supply is one of the key technologies to solve technical bottlenecks such as electrical phase separation zone, poor quality and defects in structure control algorithms traditional traction systems single co-phase schemes, an inevitable way realize development electrified railways direction safety, high speed heavy load. Based on technology, a novel quadruple scheme with negative sequence elimination suppress reactive harmonic better for two aspects system improvement algorithm...
Mapping relationships of multidimensional architectures play an essential role in the autonomous transportation system (ATS), which can help understand complex between architectures. The current mapping relationship discovery for ATS requires significant manual involvement, leading to underutilization textual data and intense subjectivity results. In order address above issues, it is necessary mine further utilize semantic information data. This study applies text-matching model...