A5055265435- Rock Mechanics and Modeling
- Parallel Computing and Optimization Techniques
- Landslides and related hazards
- Geotechnical Engineering and Analysis
- Cloud Computing and Resource Management
- Water resources management and optimization
- Distributed and Parallel Computing Systems
- Distributed systems and fault tolerance
- Water Systems and Optimization
- Water-Energy-Food Nexus Studies
- Composite Structure Analysis and Optimization
- Advancements in Battery Materials
- Geotechnical and Geomechanical Engineering
- Advanced Data Storage Technologies
- Advanced Battery Technologies Research
- Dam Engineering and Safety
- Metal and Thin Film Mechanics
- Reliability and Maintenance Optimization
- Embedded Systems Design Techniques
- Integrated Energy Systems Optimization
- Geomechanics and Mining Engineering
- Smart Cities and Technologies
- High Temperature Alloys and Creep
- Fiscal Policy and Economic Growth
- Climate change impacts on agriculture
Stony Brook University
2019-2022
Hohai University
2019-2021
State University of New York
2021
Zhejiang University
2012-2020
Chinese Academy for Environmental Planning
2020
Nanjing Tech University
2019-2020
Shenzhen Academy of Metrology and Quality Inspection
2018-2019
Peking University
2016-2017
To accurately understand the anisotropic mechanical properties of columnar jointed basalt distributed in Baihetan Hydropower Station, indoor true triaxial loading physical model tests were conducted with different joint angles (β = 0°–90°) to study strength deformation and failure mode rock mass (CJRM). The results indicate that anisotropy characteristics CJRM under stress environments are significant. peak shows a U-shaped (σ2 2 MPa, σ2 4.5 MPa), W-shaped 3 linearly increasing 5 MPa) rule...
Accurate estimation and prediction of the lithium-ion (Li-ion) batteries’ performance has important theoretical practical significance to make better use battery avoid unnecessary losses. State health (SOH) is used as a qualitative measure capability store deliver energy in system. To evaluate predict SOH batteries, Gaussian process regression with neural network (GPRNN) its variance function proposed. Experimental results confirm that proposed method can be effectively applied Li-ion...
Capacity degrading over repeated charge/discharge cycles is a main parameter for evaluating battery performance, which commonly used determining the state of health. However, it difficult to measure available capacity because requires normal operation be terminated and long time-consuming detection process. This study presents an online available-capacity estimation method by combining extended Kalman filter (EKF) with Gaussian process regression (GPR) daily partial charge data lithium-ion...
In this work, an anisotropic constitutive model of hexagonal columnar jointed rock masses is established to describe the distribution law deformation and failure joint caverns under conditions, implemented study mass at dam site Baihetan Hydropower Station on Jinsha River. The based Cosserat theory considers mesoscopic bending effect macroscopic mean. influences plane inclination equivalent elastic parameters are discussed via introduction off-axis transformation matrix analysis example. It...
Because of its special structure, the anisotropic properties columnar jointed rock mass (CJRM) are complicated, which brings difficulty to engineering construction. To comprehensively study characteristics CJRM, uniaxial compression tests were conducted on artificial CJRM specimens. Quadrangular, pentagonal and hexagonal prism models introduced, dip direction joints was considered. Based test results structural features three models, deformation strength specimens analyzed compared. The...
In this paper, we present a framework for moving compute and data between processing elements in distributed heterogeneous system. The implementation of the is based on LLVM compiler toolchain combined with UCX communication framework. can generate binary machine code or bitcode multiple CPU architectures move to remote machines while dynamically optimizing linking target platform. remotely injected recursively propagate itself other new code. goal paper threefold: (a) an ar-chitecture that...
As the number of shared-memory cores per node in modern High Performance Computing (HPC) machines continues to grow, hybrid programming models like MPI+threads are be- coming a preferred choice for scientific applications. While being able utilize computation resources efficiently, threads applications often compete with each other communication resources, resulting negative impact on performance. The OpenSHMEM distributed model provides com- munication context objects that can be used...