A5022897645- Risk and Safety Analysis
- Topic Modeling
- Occupational Health and Safety Research
- Complex Network Analysis Techniques
- Evaluation and Optimization Models
- Advanced Fiber Optic Sensors
- Infrastructure Maintenance and Monitoring
- Advanced Graph Neural Networks
- Video Analysis and Summarization
- Advanced Text Analysis Techniques
- Advanced Image and Video Retrieval Techniques
- Text and Document Classification Technologies
- Advanced Neural Network Applications
- Fault Detection and Control Systems
- Quality and Safety in Healthcare
- Spectroscopy and Chemometric Analyses
- Mental Health Research Topics
- Non-Destructive Testing Techniques
- Digital Media Forensic Detection
- Advanced Decision-Making Techniques
- Image Processing Techniques and Applications
- Complex Systems and Time Series Analysis
- Electricity Theft Detection Techniques
- Optical Coherence Tomography Applications
- Advanced battery technologies research
University of Science and Technology Liaoning
2023-2025
Shenzhen Polytechnic
2024-2025
Beijing Municipal Ecological and Environmental Monitoring Center
2021-2024
Beijing University of Posts and Telecommunications
2014-2024
Zhuhai People's Hospital
2020
Jinan University
2020
Beijing Jiaotong University
2008-2009
The heterogeneous nature of a complex network determines the roles each node in that are quite different. Mechanisms networks such as spreading dynamics, cascading reactions, and synchronization highly affected by tiny fraction so-called important nodes. Node importance ranking is thus great theoretical practical significance. Network entropy usually utilized to characterize amount information encoded structure measure structural complexity at graph level. We find can also serve local level...
Abstract The inverted perovskite solar cells based on hole‐selective self‐assembled molecules (SAMs) have been setting new efficiency benchmarks. However, the agglomeration of SAM and lack defect passivation ability are two critical issues that need to be addressed. It is demonstrated by blending co‐adsorbent 4‐phosphoricbutyl ammonium iodide (4PBAI) with 4‐(7H‐dibenzo[c,g]carbazole‐7‐yl) phosphonic acid (4PADCB), enhanced homogeneity, conductivity, better energy levels can realized for...
For self‐assembled molecule (SAM)‐based inverted perovskite solar cell, the buried interface (SAM/perovskite interface) significantly determines overall efficiency and stability of device, which requires meticulous modulation. In this work, a series phthalimide derivatives (namely 4‐(1,3‐dioxoisoindolin‐2‐yl)butan‐1‐ammonium iodide [DBAI], 2‐(1,3‐dioxoisoindolin‐2‐yl)ethan‐1‐ammonium [DEAI], 6‐(1,3‐dioxoisoindolin‐2‐yl)hexan‐1‐ammonium [DHAI]) are developed as interfacial modification...
The rapid development of computer technology has brought a large amount text information. This paper aims to classify the in view increasing number Chinese texts process supervision and demand processing texts, improve efficiency information query management. In this paper, brake classification system based on support vector machine is designed implemented. represented mathematically by space model, classifier trained principle machine. performance tested evaluated using file regulatory...
The critical aspects of risk management include hazard identification, assessment, and control. Timely is to company decision‐making, but the process acquiring knowledge often time‐consuming labor‐intensive. Knowledge graph question answering (KGQA) provides an effective solution by delivering through accurate reasoning. However, existing KGQA methods do not cover are difficult retrieve quickly accurately from large graphs. This study describes a complex method for intelligently generating...
Topological measures are crucial to describe, classify and understand complex networks. Lots of proposed characterize specific features networks, but the relationships among these remain unclear. Taking into account that pulling networks from different domains together for statistical analysis might provide incorrect conclusions, we conduct our investigation with data observed same network in form simultaneously measured time series. We synthesize a transfer entropy-based framework quantify...
Complex network methodology is very useful for complex system explorer. However, the relationships among variables in are usually not clear. Therefore, inferring association networks from their observed data has been a popular research topic. We propose synthetic method, named small-shuffle partial symbolic transfer entropy spectrum (SSPSTES), multivariate time series. The method synthesizes surrogate data, (PSTE) and Granger causality. A proper threshold selection crucial common correlation...
This paper proposes a general integration method which can effectively describe the characteristics of pipeline leakage and help distinguish multiple microstates. Since rapid development Φ-OTDR in recent years, this technology has been applied to more fields, such as fiber optic safety monitoring, seismic structural health monitoring. Among them, characteristic continuous full-scale monitoring but there are few researches on state at present. In paper, based analysis with technology,...
Fault detection in industrial process is a popular research topic. Although the distributed control system(DCS) has been introduced to monitor state of process, it still cannot satisfy all requirements for fault systems. In this paper, we proposed novel method based on topological features and support vector machine(SVM), process. The takes global information measured variables into account by complex network model predicts whether system generated some faults or not SVM. can be divided four...
In order to realize the multithreshold segmentation of images, an improved algorithm based on graph cut theory using artificial bee colony is proposed. A new weight function gray level and location pixels constructed in this paper calculate probability that each pixel belongs same region. On basis, a cost reconstructed can use both square nonsquare images. Then optimal threshold image obtained through searching for minimum value algorithm. paper, public dataset widely used images were...
Abstract The task of partial copy detection in videos aims at determine if one or more segments the query video are already present data-set, while giving information similar portion time period. At present, most effective algorithms designed as three steps: feature extraction, matching and alignment. separation alignment module ignores spatio-temporal to some extent. Therefore, satisfactory performance is not obtained. In order reduce this loss, article does decompose it into two separate...
Complex network methodology is very useful for complex system exploration. However, the relationships among variables in systems are usually not clear. Therefore, inferring association networks from their observed data has been a popular research topic. We propose method, named small-shuffle symbolic transfer entropy spectrum (SSSTES), multivariate time series. The method can solve four problems networks, i.e., strong correlation identification, quantification, direction identification and...
Complex networks provide a convenient way to model the process of occupational injury occurrence at system level, and node importance metrics are usually employed quantify influence factors leading injuries. However, traditional such as degree betweenness based on supposition that network is homogeneous one types its nodes have be same. To describe occurrence, there should least two nodes, i.e., source nodes. Since this heterogeneous in nature, for evaluation no longer applicable. Hence, we...
It is commonly known that for characteristics, such as long-distance, high-sensitivity, and full-scale monitoring, phase-sensitive optical time-domain reflectometry (Φ-OTDR) has developed rapidly in many fields, especially with the arrival of 5G. Nevertheless, there are still some problems obstructing application practical environments. First, fading effect leads to results falling into dead zone, which cannot be demodulated effectively. Second, because high sensitivity, Φ-OTDR system easy...
Road scene parsing is a common task in semantic segmentation. Its images have characteristics of containing complex context and differing greatly among targets the same category from different scales. To address these problems, we propose segmentation model combined with edge detection. We extend network an encoder-decoder structure by adding feature pyramid module, namely Edge Feature Pyramid Network (EFPNet, for short). This module uses detection operators to get boundary information then...
In recent years, with the development of artificial intelligence, power forecasting based on big data analysis has gradually become intelligent. order to improve prediction accuracy and efficiency model in dealing large volume data, this paper combines compressed sensing algorithm random forest regression algorithm. The discrete cosine transform base is used sparsely represent data. original restored by solving norm optimization problem achieve purpose denoising. And processed for...
There is a complex legal system in work safety, but currently only manual means are used to recommend law, and intelligent methods urgently needed. To address law recommendation for hard examples, we propose model using data augmentation named KeySim-FLMacBERT, which an improved version of MacBERT optimized from both algorithm perspectives: KeySim module leverages keyword library safety laws enhance the illegal fact, uses SimBERT text generation method achieve example mining; FLMacBERT...