A5016826981- Fault Detection and Control Systems
- Machine Fault Diagnosis Techniques
- Reliability and Maintenance Optimization
- Advanced Battery Technologies Research
- Network Security and Intrusion Detection
- Anomaly Detection Techniques and Applications
- Antenna Design and Optimization
- Advanced Malware Detection Techniques
- Advanced Algorithms and Applications
- Advancements in Battery Materials
- Advanced Neural Network Applications
- Radio Astronomy Observations and Technology
- Cloud Computing and Resource Management
- Advanced Data Storage Technologies
- UAV Applications and Optimization
- Video Surveillance and Tracking Methods
- Infrastructure Maintenance and Monitoring
- Complex Network Analysis Techniques
- Computational Drug Discovery Methods
- Machine Learning and ELM
- Security in Wireless Sensor Networks
- Hydraulic and Pneumatic Systems
- Photoacoustic and Ultrasonic Imaging
- Robotic Path Planning Algorithms
- Electromagnetic Compatibility and Measurements
NetEase (China)
2021-2024
University of Wollongong
2021-2024
Beijing University of Posts and Telecommunications
2024
Alibaba Group (China)
2018-2023
Beihang University
2017-2023
Nanjing University of Aeronautics and Astronautics
2019-2022
PLA Army Engineering University
2019-2021
Beijing University of Technology
2020
China Electronic Product Reliability and Environmental Test Institute
2020
Research Institute of Petroleum Exploration and Development
2020
Rotating machinery plays a key role in mechanical equipment, and the fault diagnosis of rotating is popular research topic. To overcome dependency on expert knowledge regarding conventional time-frequency analysis methods, machine learning (ML) artificial intelligence (AI)-based methods are commonly studied. Although these can achieve high-accuracy results, they based large number training samples. A generative adversarial network (GAN) an algorithm with capability generating realistic...
As combat missions become increasingly complex in both space and time, cross-regional joint operations (CRJO) is becoming an overwhelming trend modern air warfare. How to allocate resources prior the operation becomes a central issue improve efficiency. In this paper, we focus on cooperative mission planning of multiple heterogeneous unmanned aerial vehicles (UAVs) CRJO. A multi-objective optimization problem presented with aim minimizing makespan while maximizing value expectation obtained....
Nowdays, it is prevalent to train deep learning (DL) models in cloud-native platforms that actively leverage containerization and orchestration technologies for high elasticity, low flexible operation cost, many other benefits. However, also faces new challenges our work focusing on those related I/O throughput training, including complex data access with complicated performance tuning, lack of cache capacity specialized hardware match its dynamic requirement, inefficient resource sharing...
Various relations existing in Electroencephalogram (EEG) data are significant for EEG feature representation. Thus, studies on the graph-based method focus extracting relevancy between channels. The shortcoming of graph is that they only consider a single relationship electrodes, which results an incomprehensive representation and relatively low accuracy emotion recognition. In this paper, we propose fusion convolutional network (FGCN) to extract various fuse these extracted represent more...
Histotripsy has been investigated as a noninvasive, drug-free, image-guided thrombolysis method that fractionates blood clots using acoustic cavitation alone. In previous histotripsy-mediated studies, clouds were generated multi-cycle pulses and tended to form on vessel wall. To avoid potential cavitational damage the wall, new histotripsy approach, termed microtripsy, recently discovered in which is via an intrinsic-threshold mechanism single-cycle pulses. We hypothesize microtripsy can...
Photoacoustic imaging, an emerging biomedical imaging modality, holds great promise for preclinical and clinical researches. It combines the high optical contrast ultrasound resolution by converting laser excitation into ultrasonic emission. In order to generate photoacoustic signal efficiently, bulky Q-switched solid-state systems are most commonly used as sources hence limit its commercialization. As alternative, miniaturized semiconductor system has advantages of being inexpensive,...
Heap spraying is an attack technique commonly used in hijacking browsers to download and execute malicious code. In this attack, attackers first fill a large portion of the victim process's heap with Then they exploit vulnerability redirect control attackers' code on heap. Because location injected not exactly predictable, traditional heap-spraying attacks need inject huge amount executable increase chance success. Injected usually includes lots NOP-like instructions leading shellcode....
We present a fusion-oriented pulsed power module M-50, which is based on the linear transformer driver (LTD) and magnetically insulated inductive voltage adder (MIVA) technologies. The consists of 50 identical LTD cavities connected in series, one 60 modules fusion-scale facility. M-50 comprehensive test bed for integration engineering validation. Each cavity 32 bricks has an output capability $90\text{ }\text{ }\mathrm{kV}/1.0\text{ }\mathrm{MA}/120\text{ }\mathrm{ns}$ to matched load....
Abstract Histone deacetylase 3 (HDAC3) is a potential drug target for treatment of human diseases such as cancer, chronic inflammation, neurodegenerative and diabetes. Machine learning (ML) an essential cheminformatics approach has been widely used QSAR modeling. However, none them applied to HDAC3. To this end, we carefully compiled set 1098 compounds from the ChEMBL database that have assayed against HDAC3 calculated three different sets molecular features each compound, i. e....
Malware and its variants continue to pose a threat network security. Machine learning has been widely used in the field of malware classification, but some emerging studies, such as attention mechanisms, are rarely applied this field. In paper, we analyze correspondence between bytecode disassembly malware, propose new feature extraction method based on multi-dimensional sequence. Also, construct classification framework mechanism Convolutional Neural Networks mechanism. Furthermore, also...
In the design phase of Li-ion batteries for electric vehicles, battery manufacturers need to carry out cycle life tests on a large number formulations get best one that meets customer demands. However, such take considerable time and money due long power batteries. Aiming at reducing cost tests, we propose prediction method can learn historical degradation data extrapolate predict remaining trend current formulation sample taking initial stage partial test results as input. Compared with...