A5100725201- Structural Health Monitoring Techniques
- Natural Language Processing Techniques
- Probabilistic and Robust Engineering Design
- Control Systems and Identification
- Multimodal Machine Learning Applications
- Topic Modeling
- Human Pose and Action Recognition
- Structural Load-Bearing Analysis
- Infrastructure Maintenance and Monitoring
- Blockchain Technology Applications and Security
- Higher Education Learning Practices
- Financial Literacy, Pension, Retirement Analysis
- Digital and Cyber Forensics
- Distributed Control Multi-Agent Systems
- Air Traffic Management and Optimization
- Robotic Path Planning Algorithms
- Educational Technology and Assessment
- Adaptive Control of Nonlinear Systems
- Aerospace and Aviation Technology
- Sparse and Compressive Sensing Techniques
- Photoacoustic and Ultrasonic Imaging
- Surface Roughness and Optical Measurements
- Neural Networks Stability and Synchronization
- Civil and Structural Engineering Research
- Hydraulic and Pneumatic Systems
China University of Petroleum, Beijing
2025
Beihang University
2023
Civil Aviation Flight University of China
2023
University of Chinese Academy of Sciences
2020
Guangzhou Institute of Advanced Technology
2020
Chinese Academy of Sciences
2020
Shenzhen Institutes of Advanced Technology
2020
Northwestern Polytechnical University
2018-2019
Commercial Aircraft Corporation of China (China)
2019
Vanguard (United States)
2018
Automatic detection of industrial product damage using machine learning is a promising research area. At the same time, various methods based on convolutional neural networks have very important role in application visual automatic detection. Therefore, vision-based high-speed railway rail has received widespread attention. This paper proposes an efficient method for rails called SCueU-Net. For first combination U-Net graph segmentation network and saliency cues location applied to task The...
Existing end-to-end task-oriented dialog systems struggle to dynamically model long context for interactions and effectively incorporate knowledge base (KB) information into generation. To conquer these limitations, we propose a Dual Dynamic Memory Network (DDMN) multi-turn generation, which maintains two core components: memory manager KB manager. The expands the turn by keeps track of history with an updating mechanism, encourages filter irrelevant memorize important newly coming...
Image captioning is usually built on either generation-based or retrieval-based approaches. Both ways have certain strengths but suffer from their own limitations. In this paper, we propose an Interactive Dual Generative Adversarial Network (IDGAN) for image captioning, which mutually combines the and methods to learn a better ensemble. IDGAN consists of two generators discriminators, where generation- benefit each other's complementary targets that are learned dual adversarial...
As natural gas pipeline networks expand, the complexity of scheduling models increases, making feasibility analysis increasingly difficult. This study focuses on optimization for distribution network scheduling, considering it as a classification problem. Models grounded in traditional neural networks, parallel branch and graph are developed assessed. Two distinct scales explored by collecting limited dataset sample cases to train validate proposed through empirical case studies. The results...
Cross-lingual Machine Reading Comprehension (CLMRC) remains a challenging problem due to the lack of large-scale annotated datasets in low-source languages, such as Arabic, Hindi, and Vietnamese. Many previous approaches use translation data by translating from rich-source language, English, languages auxiliary supervision. However, how effectively leverage reduce impact noise introduced onerous. In this paper, we tackle challenge enhance cross-lingual transferring performance novel...
With the continuous increase in low altitude flight density, issue of navigation safety has attracted widespread attention. Due to complex environment, is more susceptible ground obstacles and weather effects than commercial aviation. In order ensure helicopters airspace, this paper proposes an improved support vector machine based conflict detection model. By modeling network areas utilizing Support Vector Machine (SVM) classification features, discrimination was achieved, ultimately...
Empirical Mode Decomposition (EMD) and Wavelet method have been widely used in the vibration signal analysis, mainly because of powerful ability blind modal separation. In flutter test, identification trajectories tracking online can help to ensure test safety. However, traditional EMD is only suitable for offline data batching, which limits its applications real-time or near analysis. For this reason, according processing requirements under turbulent excitation conditions, combining with...
Improving financial and health literacy is an important step in reducing economic vulnerability older age, yet the means by which individuals accumulate these types of human capital remains open question. This paper evaluates impact online search activities on levels literacy. We find that using internet for such information increases significantly: doing so frequently (versus not at all) 16 percent, 12 percent. Our results are robust to alternative measures They also instrumental variable...
Because of the growth Internet, everyone uses a lot applications and web services. Each user needs an account when using network service or application. Companies businesses also link accounts with information, resources, permissions. Therefore, security is paramount importance. This article discusses how to protect users' prevent problems caused by disclosure some their personal information. A new method storing authorized information proposed improve account.
The quadrotor Unmanned Aerial Vehicle (UAV) possesses various advantages, such as miniaturization, full autonomy, and a simple structure. However, its attitude control stability is challenging to ensure due inherent characteristics, including underactuation, high coupling, nonlinearity, susceptibility disturbances. To address these issues, this paper conducts dynamic modeling of the quadrotor's actuators, ring, position followed by controller design. A comparative study performed between...
In this paper, a robust H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">∞</sub> filter for networked control system is designed. A continuous model of investigated and it denoted by the one that similar to time-delay system, through some assumption transformation. Because only limited information can be transported in network, time-invariant logarithmic quantization considered output system. The filtering error designed paper robustly...
The analysis on the test data is an important way to obtain flutter characteristics of aircraft. Because low quality measured signal, preprocessing indispensable improve accuracy analysis. In this paper, morphological filtering introduced into field structure processing. theory mathematical morphology given, and effectiveness depending different operation types structural elements are discussed. After that, typical boundary prediction (FBP) method used verify validity feasibility through...
The flutter test with progression variable speed is actively explored in recent years. This paper proposes an improved Kalman smoothing filter (EM-KS) algorithm based on expectation maximization for the non-stationary characteristics of signal this type experiment, which can effectively improve estimation accuracy time-varying parameter modeling. Combining time domain criterion, a new method boundary prediction that be recursively implemented given. Finally, reliability and engineering...