A5102823841- Advanced machining processes and optimization
- Industrial Technology and Control Systems
- Multimodal Machine Learning Applications
- Advanced Sensor and Control Systems
- Advanced Machining and Optimization Techniques
- Machine Fault Diagnosis Techniques
- Gear and Bearing Dynamics Analysis
- Advanced Algorithms and Applications
- Electric and Hybrid Vehicle Technologies
- Electromagnetic Launch and Propulsion Technology
- Heat Transfer and Optimization
- Laser and Thermal Forming Techniques
- Video Analysis and Summarization
- Energetic Materials and Combustion
- Metallurgy and Material Forming
- Robot Manipulation and Learning
- Quantum Information and Cryptography
- Advanced Computational Techniques and Applications
- Quantum and electron transport phenomena
- Quantum Computing Algorithms and Architecture
- Advanced Surface Polishing Techniques
- Manufacturing Process and Optimization
- Human Pose and Action Recognition
- Hydraulic and Pneumatic Systems
- Guidance and Control Systems
Westlake University
2024
Xi'an Technological University
2006-2024
Zhejiang University
2024
Beijing Institute of Technology
2023
Xi'an University of Architecture and Technology
2022
Beijing Academy of Quantum Information Sciences
2021
University of Science and Technology of China
2021
Qingdao University of Science and Technology
2010-2015
Florida State University
2008-2014
Hefei University of Technology
2010
Scaling up to a large number of qubits with high-precision control is essential in the demonstrations quantum computational advantage exponentially outpace classical hardware and algorithmic improvements. Here, we develop two-dimensional programmable superconducting processor, Zuchongzhi, which composed 66 functional tunable coupling architecture. To characterize performance whole system, perform random circuits sampling for benchmarking, system size 56 20 cycles. The cost simulation this...
Vision-language-action models (VLAs) have become increasingly popular in robot manipulation for their end-to-end design and remarkable performance. However, existing VLAs rely heavily on vision-language (VLMs) that only support text-based instructions, neglecting the more natural speech modality human-robot interaction. Traditional integration methods usually involves a separate recognition system, which complicates model introduces error propagation. Moreover, transcription procedure would...
Using the CFX software and Large Eddy Simulaion (LES) method, this paper numerically simulates hydrodynamic characteristics of flow around a 3d circular cylinder at high Reynolds number (Re=5.6×10 3 , 2.8×10 4 1.1×10 5 ). The numerical simulation focuses on investigating vortex shedding angle, tube, base pressure, static fluctuating drag lift. result calculation shows that forces every section along span are symmetrical with respect to middle section. Moreover obviously appears three...
Wear debris generated in a pitting process of planetary gearboxes carries valuable information about health status. However, RGB images online wear are often affected by image blur caused Gaussian noise, high-frequency and other random noises besides including bubbles lubricating oil, dark oil contamination, the temperature rise electronic components. To address these issues, this work, an denoising model WVBOD was proposed based on fusion wavelet, variational mode decomposition non-local...
An electromagnetic railgun is one of the applications launchers, which are devices used to accelerate projectiles velocities exceeding those attained with conventional propelling systems. A magnetic field generated when current flows through two conductive rails connected by a moving armature. The that passes exerts an force on armature and causes it high speeds. In this work, three-dimensional transient model for thermal solutions launch process presented. accounts determination temperature...
A numerical method to study the effect of internal heat sink variables on tool-chip interface temperature in orthogonal cutting was presented. The analytical is based two main assumptions that chip can be treated as a semi-infinite body with moving source at and tool stationary uniform plane inside body. An approach using point partition coefficient employed obtain distribution. distributions along different intensities, distances from interface, areas were effects intensity, distance area...
When measuring the coaxiality of a cylinder, we often get poor reproducibility measurement results. This phenomenon is related not only to its machine condition, but also method. The main factors that cause errors are analyzed in this paper, and some solutions improve results provided.
To ensure a long-term quantum computational advantage, the hardware should be upgraded to withstand competition of continuously improved classical algorithms and hardwares. Here, we demonstrate superconducting computing systems \textit{Zuchongzhi} 2.1, which has 66 qubits in two-dimensional array tunable coupler architecture. The readout fidelity 2.1 is considerably an average 97.74\%. more powerful processor enables us achieve larger-scale random circuit sampling, with system scale up 60 24...
Under high speeds and heavy loads, gear tooth surfaces will be worn, which lead to accidents. The current wear prediction model is not comprehensive, but the that considers temperature field dynamic response can provide more accurate fault early warning than previous methods. Therefore, according Archard model, discrete of surface deduced in this paper, evolution trend predicted analyzed each operation stage. was simulated by finite-element method, improved considering its influence on...
Fueled by the Large Language Models (LLMs) wave, Visual-Language (LVLMs) have emerged as a pivotal advancement, bridging gap between image and text. However, video making it challenging for LVLMs to perform adequately due complexity of relationship language spatial-temporal data structure. Recent Video-Language (LVidLMs) align feature static visual like into latent space feature, general multi-modal tasks leverage abilities LLMs sufficiently. In this paper, we explore fine-grained alignment...
This paper addresses the inherent inference latency challenges associated with deploying multimodal large language models (MLLM) in quadruped vision-language-action (QUAR-VLA) tasks. Our investigation reveals that conventional parameter reduction techniques ultimately impair performance of foundation model during action instruction tuning phase, making them unsuitable for this purpose. We introduce a novel latency-free MLLM model, dubbed QUART-Online, designed to enhance efficiency without...
Regenerative braking of electric vehicles can not only conserve energy, but also increase driving range. control strategy formulate proportional relation regenerative and mechanical braking, how to distribution has become the key problem raising energy[1]．This paper research on force in process, mathematical models put forward a new strategy. The is simulated analyzed ADVISOR. Used strategy, results indicate that fuel economy, emission dynamic simulation vehicle are improved.
When we use Coordinate Measuring Machine to measure some points of the workpieces, wrong measuring method will get poor repeatability and reproducibility measurement results. Now vector CMM is a more convenient way. The main principles are analyzed in this paper, workpiece measured for example.
Based on the importance of regenerative braking system (RBS) to EV’s sustained travel mileage and prolonging life mechanical brake. This paper introduces characteristic principle system, at same time analyses pattern brake energy recovery, emphasize distribution force. At last, this provides good base for developing specific control strategy.