A5101523284- Meteorological Phenomena and Simulations
- Smart Grid and Power Systems
- Geophysics and Gravity Measurements
- Parallel Computing and Optimization Techniques
- Computational Physics and Python Applications
- Neural Networks and Reservoir Computing
- Nonlinear Waves and Solitons
- Photonic and Optical Devices
- Advanced Adaptive Filtering Techniques
- Elasticity and Material Modeling
- Carbohydrate Chemistry and Synthesis
- Distributed systems and fault tolerance
- Aquatic and Environmental Studies
- Image Enhancement Techniques
- Blind Source Separation Techniques
- Error Correcting Code Techniques
- Advanced Measurement and Detection Methods
- Chemical Synthesis and Reactions
- Photonic Crystals and Applications
- Aerospace and Aviation Technology
- Image and Object Detection Techniques
- Distributed and Parallel Computing Systems
- Advanced Numerical Analysis Techniques
- Advanced Image Processing Techniques
- Data Visualization and Analytics
Tsinghua University
2013-2024
Beihang University
2005-2022
Anhui University of Technology
2022
National Supercomputing Center in Wuxi
2019
State Grid Corporation of China (China)
2016
Beijing University of Chemical Technology
2012
Beijing Institute of Technology
2007
Hong Kong Baptist University
1994-2003
Abstract. Atmosphere modelling applications are becoming increasingly memory-bound due to the inconsistent development rates between processor speeds and memory bandwidth. In this study, we mitigate bottlenecks reduce computational load of Global–Regional Integrated Forecast System (GRIST) dynamical core by adopting a mixed-precision computing strategy. Guided an application iterative principle, identify coded equation terms that precision insensitive modify them from double single...
We propose a novel broad-angle polarization beam splitter utilizing the spatial dispersion of special multi-layered dielectric periodic structure. The equal-frequency contours this structure are flat lines for TE but curved TM at designed frequency. This has fixed optical thickness all incident angles. A working over broad range angle (from 0 degrees to 70 ) is achieved by stacking two such structures finite length (in normal direction) with half-period shift in transverse direction.
Emotional classification is an important task in natural language processing. Aiming at the problem of gradient explosion and disappearance traditional cyclic neural network text affective analysis model, high computational cost, this paper proposes a bi-directional GRU model based on attention mechanism for analysis. Corp-Htl-ba-6000 dataset trained tested, good results are obtained, which can effectively improve effect emotional speed.
Summary form only given. As the method to study long-term climate trend and predict potential risk, modeling is becoming a key research topic among governments organizations. One of most essential challenging components in atmospheric model. To cover high resolution simulation scenarios, developers have face challenges from billions mesh points extremely complex algorithms. Shallow Water Equations (SWEs) are set conservation laws that perform characteristics atmosphere. The SWEs can serve as...
This article presents Jdebug, a fast, non-intrusive and scalable fault locating tool for extreme-scale parallel applications. Large-scale debugging has drawn more attention with the increasing scale of supercomputers To eliminate program intrusion caused by traditional instrumentation or interception during information acquisition, we introduce out-of-band management into large-scale debugging. We propose rapid gathering scheme that separates user traffic to solve scalability problem...
Fault-phase selection is an important link on corrective actions of high-voltage transmission lines protection, but also prerequisite for accurate fault location.The traditional phase method only applicable to simple fault, low sensitivity, and easily affected by system operation mode, grounding resistance etc.. On the basis analysis component each current, this paper gives a definition energy concept signal presents new type identification based coefficient.This directly uses current...
In this research, a high performance computing weather forecasting application GRAPES has been tuned onto functional unit (FU) array based architecture. Software and hardware approaches are specifically employed to increase the data locality reuse accelerate stencil computation in GRAPES. The simulation results indicate that we can achieve per-core average IPC of 12.3 within 20-stage FU processor, which 5.8x power-efficiency boost than many-core processor (MCP) same process technology. This...
Based on the former work about variable upper limit integral of Bézier curve, construction set made by all degree n curves will be studied. Furthermore, for different n, sets corresponding could build a like pyramid.
Butterfly optimization algorithm (BOA) is a new swarm intelligence mimicking the behaviors of butterflies. However, there still much room for improvement. In order to enhance convergence speed and accuracy BOA, we present an improved SCLBOA based on SIBOA, which incorporates logical mapping Lévy flight mechanism. The chaotic map used population initialization, then mechanism integrated into algorithm. To evaluate performance SCLBOA, conducted many experiments standard test functions....