A5103115726- Railway Engineering and Dynamics
- Radar Systems and Signal Processing
- Aerodynamics and Fluid Dynamics Research
- PAPR reduction in OFDM
- Structural Engineering and Vibration Analysis
- Advanced Neural Network Applications
- Vibration and Dynamic Analysis
- Fluid Dynamics and Vibration Analysis
- Civil and Geotechnical Engineering Research
- Advanced Image and Video Retrieval Techniques
- Hydraulic and Pneumatic Systems
- Energy Load and Power Forecasting
- Reinforcement Learning in Robotics
- AI-based Problem Solving and Planning
- CCD and CMOS Imaging Sensors
- Domain Adaptation and Few-Shot Learning
- Optical Network Technologies
- Robotics and Sensor-Based Localization
- Cerebral Palsy and Movement Disorders
- Iterative Learning Control Systems
- Direction-of-Arrival Estimation Techniques
- Anomaly Detection Techniques and Applications
- Time Series Analysis and Forecasting
- Robotics and Automated Systems
- Prosthetics and Rehabilitation Robotics
Harbin Institute of Technology
2019-2024
Beijing Jiaotong University
2024
Ningbo University
2023-2024
Southwest Jiaotong University
2023-2024
Zhejiang Sci-Tech University
2024
Chang'an University
2023
Carnegie Mellon University
2023
University of Leeds
2023
South China University of Technology
2021
Chemotherapy Foundation
2017
Deep learning has had remarkable success in robotic perception, but its data-centric nature suffers when it comes to generalizing ever-changing environments. By contrast, physics-based optimization generalizes better, does not perform as well complicated tasks due the lack of high-level semantic information and reliance on manual parametric tuning. To take advantage these two complementary worlds, we present PyPose: a robotics-oriented, PyTorch-based library that combines deep perceptual...
Large language models (LLMs) have shown the potential to be integrated into human daily lives. Therefore, user preference is most critical criterion for assessing LLMs' performance in real-world scenarios. However, existing benchmarks mainly focus on measuring models' accuracy using multi-choice questions, which limits understanding of their capabilities real applications. We fill this gap by proposing a comprehensive Chinese benchmark SuperCLUE, named after another popular LLM CLUE....
The combination of self-supervised learning and adversarial training (AT) can significantly improve the robustness models. However, (self-AT) still lags behind that state-of-the-art (SOTA) supervised AT (sup-AT), even though performance current models has already matched or surpassed SOTA This issue raises concerns about secure application inclusion turns self-AT into a challenging joint optimization problem, recent studies have shown data augmentation methods necessary for constructing...
Bridge time-varying deformation is the time-spatial dynamic of a bridge excited by train motion, which serves as direct alignment affecting train's response and usually moves ahead train. In this study, we investigated influence on vertical acceleration with 16 cars long-span suspension through train-track-bridge coupled vibration analysis chord measurements spatial alignment. The reveal mechanism different positions car-body cars. Furthermore, developed simple method to identify most...
The nonlinear aeroelastic behavior of a folding fin in supersonic flow is investigated this paper. finite element model the established and deployable hinges are represented by three torsion springs with freeplay nonlinearity. aerodynamic grid point assumed to be at center each box for simplicity. governing equation given using infinite plate spline method modified linear piston theory. An improved fixed-interface modal synthesis method, which can reduce rigid connections interface,...
When railways traverse densely populated industrial areas, the periodic operation of large-scale factory equipment can produce consistent and regular environmental excitation on bridge structures. This excitation, if resonant with structure, may cause significant vibration bridge, posing a threat to trains’ running safety. For first time, this paper tackles practical concern by combining field measurements an analysis how environment-induced vibrations impact operational performance...
In gas sensor networks, users can access the data collected by nodes, but there is a risk of leakage during transmission. This paper proposes lightweight bidirectional authentication protocol based on physically unclonable functions (GS-PUFs) with technology to guarantee reliability from nodes. A PUF array constructed preparing sensors enhance security physical layer and reduce hardware resource consumption. The part mainly uses encryption methods, consisting data, one-way cryptographic hash...
This study examined the influence of nonuniform wind fields in valleys on dynamic responses high-speed trains and long-span bridges. The field a valley is considerably different from theoretical uniform terms mean fluctuating characteristics because speed-up effect crosswinds. spatial nonuniformity along bridge span equivalent to "sudden change" load for running trains, which increases complexity train–bridge system may diminish performance trains. To better understand valleys, multipoint...
Due to the spectrum and complexity constraints, integrated radar communication (RadCom) systems have been widely studied recently. This paper considers a typical vehicle-to-everything (V2X) scenario with frequency-sensitive targets in frequency-selective channels. Thus, transmit power on each subcarrier needs be allocated wisely so as guarantee both performance. We propose an adaptive allocation scheme two stages. In target detection stage, channel capacity probability are jointly optimized,...
The moisture content of fabrics directly affects the physical properties fabrics, it is important to accurately detect fabrics. To verify feasibility hyperspectral imaging technology for detecting cotton and thus improve accuracy detection, 100 groups images with different thicknesses contents were acquired, characteristics spectral curves elaborated, effects thickness on analyzed. Based research above, 1190, 1450 1940 nm bands determined as characteristic bands, gradient method was selected...
Data-driven methods such as reinforcement and imitation learning have achieved remarkable success in robot autonomy. However, their data-centric nature still hinders them from generalizing well to ever-changing environments. Moreover, collecting large datasets for robotic tasks is often impractical expensive. To overcome these challenges, we introduce a new self-supervised neural-symbolic (NeSy) computational framework, imperative (IL), autonomy, leveraging the generalization abilities of...
Abstract Due to the spectrum and complexity efficiency, integrated radar communications (RadCom) systems have been widely favored, in which orthogonal frequency division multiplexing (OFDM) is most popular signal conduct two functions simultaneously. However, an unoptimized pulse could suffer from severe inter-carrier interference (ICI) high out-of-band emission (OOBE), greatly degrades system performance. In this paper, we introduce shaping scheme dedicated RadCom systems, both transmitter...
Trajectory planning in robotics aims to generate collision-free pose sequences that can be reliably executed. Recently, vision-to-planning systems have garnered increasing attention for their efficiency and ability interpret adapt surrounding environments. However, traditional modular suffer from increased latency error propagation, while purely data-driven approaches often overlook the robot's kinematic constraints. This oversight leads discrepancies between planned trajectories those are...
Large language models have already demonstrated their formidable capabilities in general domains, ushering a revolutionary transformation. However, exploring and exploiting the extensive knowledge of these to comprehend multi-omics biology remains underexplored. To fill this research gap, we first introduce Biology-Instructions, large-scale biological sequences-related instruction-tuning dataset including DNA, RNA, proteins, multi-molecules, designed bridge gap between large (LLMs) complex...
The influence of the freeplay on ground vibration responses folding fin and nonlinear modelling freeplay, which exists in joints connecting inboard outboard fin, are studied this paper. first five modal parameters obtained by hammer tests linear structure without freeplay. sweep base excitation is then applied to root-fixed with adjustable experimental results show that existence can lead some phenomena structure, such as a forward backward difference, jump, multi harmonics, frequency shift....
Bridge sensors are widely used in military and civilian fields, their demand gradually increases each year. Digital the fields. High-precision low-power analog-to-digital converters (ADCs) as sensor read-out circuits a research hotspot. Sigma-delta ADC based on switched-capacitor topology have advantages of high signal-to-noise ratio (SNR), good linearity, better compatibility with CMOS processes. In this work, fourth-order feed-forward sigma-delta modulator digital decimation filter...
Neural decoding from spiking activity is an essential tool for understanding the information encoded in population neurons, especially applications like brain-computer interface (BCI). Various quantitative methods have been proposed and shown superiorities under different scenarios respectively. From machine learning perspective, task to map high-dimensional spatial & temporal neuronal low-dimensional physical quantities (e.g., velocity, position). Because of complex interactions...