A5027931296- Particle accelerators and beam dynamics
- Particle Accelerators and Free-Electron Lasers
- Magnetic confinement fusion research
- Privacy-Preserving Technologies in Data
- Reinforcement Learning in Robotics
- Stochastic Gradient Optimization Techniques
- Robotics and Sensor-Based Localization
- Speech and Audio Processing
- Tactile and Sensory Interactions
- IoT and Edge/Fog Computing
- Age of Information Optimization
- Laser-Plasma Interactions and Diagnostics
- Non-Invasive Vital Sign Monitoring
- Infrastructure Maintenance and Monitoring
- Cloud Computing and Resource Management
- Superconducting Materials and Applications
- Healthcare Technology and Patient Monitoring
- Digital Filter Design and Implementation
- Image and Object Detection Techniques
- Spam and Phishing Detection
- Mobile Crowdsensing and Crowdsourcing
- Smart Grid Security and Resilience
- Direction-of-Arrival Estimation Techniques
- Data Stream Mining Techniques
- Hydrological Forecasting Using AI
Hohai University
2017-2025
Ministry of Water Resources of the People's Republic of China
2021-2025
State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering
2022
Institute of Modern Physics
2018
Chinese Academy of Sciences
2018
University of Wisconsin–Madison
2006-2017
Madison Group (United States)
2007
South China University of Technology
2004
Abstract Digital twin requires establishing a self‐updated model to simulate the structural damage perceived onsite. Despite great success in identification and quantification, difficulty registration still limits efficiency of updating. This study presented framework that enables finite element (FE) welded joints remesh itself for updating geometric changes caused by fatigue crack. Leveraging linear geometry weld, crack algorithm was proposed automation perception. First, dual‐task network...
Health and safety management has been an important issue in construction industry. National regulations impose the using of hard hats sites. However, there are often cases which workers neglect regulations. It is desired to monitor correct wearing hat real time explore monitoring techniques facilitated by deep-learning algorithms. In this paper, a convolutional neural network based hard-hat detection algorithm proposed. algorithm, assisted computer vision technique where deep learning model...
Guidance systems for visually impaired persons have become a popular topic in recent years. Existing guidance on the market typically utilize auxiliary tools and methods such as GPS, UWB, or simple white cane that exploits user’s single tactile auditory sense. These methodologies can be inadequate complex indoor environment. This paper proposes multi-sensory system provide advice using ORB-SLAM YOLO techniques. Based an RGB-D camera, local obstacle avoidance is realized at level through...
The convergence of artificial intelligence and the Internet Things (IoT) has made remarkable strides in realm industry. In context AIoT edge computing, where IoT devices collect data from diverse sources send them for real-time processing at servers, existing message queue systems face challenges adapting to changing system conditions, such as fluctuations number devices, size, frequency. This necessitates development an approach that can effectively decouple handle workload variations...
Abstract Multi‐agent reinforcement learning relies on reward signals to guide the policy networks of individual agents. However, in high‐dimensional continuous spaces, non‐stationary environment can provide outdated experiences that hinder convergence, resulting ineffective training performance for multi‐agent systems. To tackle this issue, a novel scheme, Mutual Information Oriented Deep Skill Chaining (MioDSC), is proposed generates an optimised cooperative by incorporating intrinsic...
Stochastic computing (SC) is a low-cost computation paradigm that can replace conventional binary arithmetic to provide low hardware footprint with high scalability. However, since the SC bitstream length grows precision of represented data, regardless its lower power consumption, convolutional SC-based neural networks may not be efficient in area and energy. This work proposes novel accelerator, PSC-Conv, implement layer using new binary-interfaced stochastic architecture. PSC-Conv exploits...
This article presents the design and evaluation of an engagement-free contactless vital signs occupancy monitoring system called BedDot. While many existing works demonstrated estimation, they do not address practical challenge environment noises, online bed detection, data quality assessment in real-world environment. work a robust signal algorithm consisting three parts: 1) detection; 2) movement 3) heartbeat to identify high-quality data. It also series innovative estimation algorithms...
Blind guidance system has always been a research hotspot for years. Although there are many kinds of blind systems on the market, most them prompt from perspective single sense tactile or auditory. The method can be unstable and it does not fully mobilize other general senses with vision impairment. This paper designs implements multi-sensory that provides auditory sensations by using ORB-SLAM YOLO techniques. Based RGB-D camera, local obstacle avoidance is realized at level through point...
The proton accelerator of the China Initiative Accelerator Driven Subcritical System (CiADS) adopts continuous wave (CW) and superconducting technical route, it may accelerate 5 mA beam to 500 MeV in energy. In this paper, baseline physics design section (SC) some choices are discussed, error study results also presented.
The continual improvement of semiconductor technology has provided rapid advancements in device frequency and density. Designers electronics systems for high-energy physics (HEP) have benefited from these advancements, transitioning many designs fixed-function ASICs to more flexible FPGA-based platforms. Today’s FPGA devices provide a significantly higher amount resources than those available during the initial Large Hadron Collider design phase. To take advantage capabilities future FPGAs...
Image scrambling is an important technology in information hiding, where the Arnold transformation widely used.Several researchers have proposed application of Hénon map square image scrambling, and certain improved technologies require many times to achieve a good effect without resisting chosen-plaintext attack although it can be directly applied non-square images.This paper presents algorithm, which resist based on chaotic two-dimensional Sine Logistic modulation (2D-SLHM).Theoretical...
It is interesting to detect stator inter-turn fault, which can cause phase ground or short circuits in the stator. First, two major harmonics are found rotor current of doubly-fed Induction Generator with fault through electromagnetic analysis. By building model doubly fed induction generator based on finite element analysis, simulation result shows magnitude harmonic component at frequency (2-s)f increases more significantly fault. So it sensitive verified experimental results.
A prototype of an analog, transverse feedback system for damping the two-stream (e-p) instability has been developed and successfully tested at Los Alamos Proton Storage Ring (PSR). This paper describes configuration, results several experimental tests studies optimization along with factors limiting its performance.
The Spallation Neutron Source (SNS) at the Oak Ridge National Laboratory is a high-intensity proton-based accelerator that produces neutron beams for neutron-scattering research. As most powerful pulsed source in world, SNS has experienced an unprecedented beam instability wide bandwidth (0 to 300 MHz) and fast growth time (10 100 μs). In this paper, we propose analyze several FPGA-based designs active feedback damping system. This signal processing system first design of wideband...
For the guarantee of long-distance transport bunches China Initiative Accelerator Driven System (CIADS), a new scheme is proposed that extra magnetic field used in accelerator-target coupling section before windowless target to minimize self-modulation (SM) mechanism. Particle-in-cell simulations are carried out study influence solenoidal on mechanism when long proton move background plasmas. The similar these linear accelerator CIADS. It found presence will significantly inhibit process....
A field-programmable gate array (FPGA)-based transverse feedback damper system was implemented in the Spallation Neutron Source (SNS) accumulator ring with intention to stabilize electron–proton (e–p) instability frequency range of 1–300 MHz. The could also be used as a diagnostic tool by measuring beam transfer function (BTF). An analysis BTF measurements provides stability diagram for production at SNS. This paper describes and its setup tool. Experimental results are presented is analyzed...
Abstract Adaptive beamforming algorithms often suffer from performance degradation due to the presence of desired signal component in interference-plus-noise covariance matrix (IPNCM) and mismatch steering vector. In order improve robustness algorithm performance, a new robust adaptive is proposed this paper. algorithm, Capon spatial spectrum first used reconstruct IPNCM. Then, based on vector uncertain set, constraints are added so that does not converge interference direction. The...