A5103028012- Hydraulic and Pneumatic Systems
- Cavitation Phenomena in Pumps
- Vehicle License Plate Recognition
- Handwritten Text Recognition Techniques
- Wireless Networks and Protocols
- Video Surveillance and Tracking Methods
- High voltage insulation and dielectric phenomena
- Hand Gesture Recognition Systems
- Smart Parking Systems Research
- Advanced Sensor and Control Systems
- Advanced Image and Video Retrieval Techniques
- Image Processing Techniques and Applications
- Advanced Measurement and Detection Methods
- Advanced Vision and Imaging
- Wireless Communication Networks Research
- Advanced Steganography and Watermarking Techniques
- Embedded Systems and FPGA Design
- Brain Tumor Detection and Classification
- Engineering Diagnostics and Reliability
- Advanced Algorithms and Applications
- Oil and Gas Production Techniques
- Thermal Analysis in Power Transmission
- Machine Fault Diagnosis Techniques
- Coal Combustion and Slurry Processing
- Cell Image Analysis Techniques
Chengdu University of Traditional Chinese Medicine
2025
University of Science and Technology Beijing
2023-2025
Sun Yat-sen Memorial Hospital
2024
Sun Yat-sen University
2024
Peking University
2008-2023
Shenyang Institute of Computing Technology (China)
2023
Chinese Academy of Sciences
2023
Southwest University of Science and Technology
2023
Pingdingshan University
2023
University of Chinese Academy of Sciences
2023
Moving object detection is very important in intelligent surveillance. In this paper, an improved algorithm based on frame difference and edge presented for moving detection. First of all, it detects the edges each two continuous frames by Canny detector gets between images. And then, divides image into several small blocks decides if they are areas comparing number non-zero pixels to a threshold. At last, does block-connected component labeling get smallest rectangle that contains object....
Planetary gearbox vibrations exhibit complex amplitude modulation (AM) and frequency (FM) characteristics. The spectral correlation (SC) can reveal the cyclostationarity of rotating machinery signals, but previous studies have modeled signals as impulse response rather than AM–FM signals. This paper presents a fault diagnosis method for planetary gearboxes via SC based on model. Firstly, theoretical expression characteristics AM-FM is derived, showing that their demodulation features consist...
Pressure and vibration displacement value are relatively measured by 14 pressure sensors 2 distributing inside the tank-type model axial flow pump device under different rates. By comparison, it is found that pulsation on inlet of impeller main cause hydraulic induced device, to have similar amplitude trend with vertical as rates increases large correlation coefficient horizontal high through time-domain analysis. frequency-domain analysis, frequency three multiplies shaft frequency, but one...
The inlet flow conditions will directly affect impeller performance, which is of great concern to pump designers. In this study, based on two axial-flow devices, the influence evaluation criteria and numerical grid scales accuracy simulation are investigated, correctness verified by experiments. axial velocity distribution uniformity, weighted average angle hydraulic loss calculated with three commonly used in engineering. applicability turbulence models engineering verified. uniformity...
RFID is a popular automatic identification technology currently, which large opportunities. Active system has more advantages in long distance applications than passive system, but it also face of the collision problem when many cards work together. On basis comparing today's mainly used anti-collision algorithms, an improved algorithm (DCMA) proposed for active this paper. The designed and simulated, finally tested hardware platform. simulation shows low power advantage DCMA
As researches on robots and robotic arms going further, human portrait painting robot appears as a typical art creation application. Traditional ways to achieve this goal mainly depend the geometric texture features with no semantic information. In system, enrich portrait, key points of faces are fused gray scale During process drawing, different components can be distinguished such hair or eyes by strokes. Besides, sketch pattern is added line drawing model using cluster algorithm presents...
For wireless mesh networks (WMN), energy balance has a significant impact on network lifetime. With the rapid growth of intelligent devices, learning from environment and experience is challenge for conventional routing protocols. To alleviate this problem, paper proposes an balanced protocol based Q-learning (QEBR) WMN. Routing decisions are made by distributed reinforcement models. A concept neighbor sorting proposed in QEBR to design state reward function model. Simulation results...
Fall ranks first among the elderly aged 65 and above. In order to detect falls of living alone, we use a vision-based detection method complete fall for on low-cost small mobile robot. We propose deep learning framework this framework, Raspberry Pi 4 Model B is selected as hardware platform robot, lightweight NanoDet-Lite used detection. The mAP our 0.912 model size only 2.17MB. Our works more than 3 times faster YOLOv3-tiny without any accelerator. ncnn at 22.03 FPS reaches 0.902. results...
Considering its advantages in dealing with high-dimensional visual input and learning control policies discrete domain, Deep Q Network (DQN) could be an alternative method of traditional auto-focus means the future. In this paper, based on Reinforcement Learning, we propose end-to-end approach that can learn from finish at a clear spot automatically. We demonstrate our - discretizing action space coarse to fine steps applying DQN is not only solution but also general towards vision-based...
As the functions of hydraulic turbine pump including self-adjusting and compensation with each other, it is far-reaching to analyze its internal flow by numerical simulation based on CFD, mainly pressure field velocity in pump.The three-dimensional models are made Pro/Engineer software;the fields simulated numerically CFX ANSYS software. According results design condition, analyzed respectively .The findings show that static decreases systematically gradient obvious area turbine; increases...
In this article, the surface charge distribution and flashover characteristics of epoxy basin insulators under ac–dc composite voltage with specific proportions in a down-sized dc-gas-insulated switchgear (GIS) are investigated by experiments. Results show that more heteropolar charges observed on insulator tend to diffuse toward central high-voltage electrode increasing amplitude. Higher prestressing would result decline ac component but lifting absolute amplitude voltage. With direct ramp...
Abstract The stability problem for the uncertain time‐varying delayed neutral‐type system is concerned in this paper. By introducing a novel Lyapunov‐Krasovskii functional (LKF) related to delay‐product‐type function and two delay‐dependent matrices, some new robust sufficient conditions are derived, which based on convex linear matrix inequality (LMI) framework. paper can reduce conservativeness of recent previous ones. In end, numerical examples, including system, partial element...
For heterogeneous networks (HetNets), load balance and energy have a significant impact on cell range expansion (CRE) network lifetime extension (NLE). With the rapid growth of intelligent devices, obtaining optimal bias values routing targets for mobile user equipments (UEs) in HetNets can be challenging by using conventional selection strategies. To alleviate this problem, paper combines two Q-learning based strategies (QSS) to relief imbalance imbalance. Here, decisions are made...
Abstract In this paper, we propose a license plate recognition model, which can detect and recognize the in single forward. The features of input image are extracted by our 15-layer convolutional neural network. detection branch, use loss function with better nonlinear to fit process plate. To catch location less information loss, add Intersection over Ground-truth (IoG) into Union (IoU) get Balanced-IoU (BIoU loss). combination these two functions make model predictive result. introduce an...
Nanostructured energetic materials have attracted considerable research interests during the past decades because of their improved performances in thermal decomposition and combustion. In this work, a porous nanosheet structure dihydroxylammonium 5, 5′-bistetrazole-1, 1′-diolate (TKX-50) has been fabricated by facile ice templating strategy, which is based on self-assembly TKX-50 rapid recrystallization. Thermal properties were determined differential scanning calorimetry/thermogravimetry...
The large deviations between the partial pressure of oxygen and hydrogen can lead to severe membrane damage, which reduces life polymer electrolyte fuel cell (PEMFC). deviation be restrained by effective control pressure. Considering actual operating environment PEMFC, its system is not only affected load changes, but also some unknown disturbances, such as parameter variations, failures, etc. above factors, this paper has proposed a robust nonlinear adaptive (RNAPC) employing perturbation...
The nursing care of bed-ridden patients is an overworked task, in which the attitude testing bed-assistants important technique. At present, there are three main approaches to detect a person's posture on bed, i.e., pressure sensor arrays, wearable devices, and image recognition methods. While, method faced with lack precision high cost challenge, devices face problem comfort, methods have risk violating privacy. In this paper, new based passive multi-label RFID array mattress system...
Abstract Recent advancements in cancer immunotherapy have improved patient outcomes, yet responses to remain moderate. We conducted a Phase II clinical trial ( NCT04718415 ) involving 51 patients undergoing neoadjuvant chemoimmunotherapy and applied single-cell RNA T/BCR sequencing on tumor blood samples elucidate the immune cell perturbations. Our findings associate poor response with reduced levels of CCR7 + CD4 Naïve T cells CD27 Memory B cells, as well higher expression...
Motivation: Non-Gaussian diffusion models can effectively characterize the microstructure of tissues. Goal(s): To investigate potential predictive value multiple non-Gaussian for assessing cervical cancer (CC). Approach: Diffusion parameters derived from continuous-time random-walk (CTRW), diffusion-kurtosis imaging (DKI), fractional order calculus (FROC) and intravoxel incoherent motion (IVIM) were calculated. The most significant histogram features selected by univariate analysis...
Brain tumor segmentation models have aided diagnosis in recent years. However, they face MRI complexity and variability challenges, including irregular shapes unclear boundaries, leading to noise, misclassification, incomplete segmentation, thereby limiting accuracy. To address these issues, we adhere an outstanding Convolutional Neural Networks (CNNs) design paradigm propose a novel network named A4-Unet. In A4-Unet, Deformable Large Kernel Attention (DLKA) is incorporated the encoder,...