X. He

ORCID: 0009-0008-1202-7784
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Advanced DC-DC Converters
  • Multilevel Inverters and Converters
  • Structural Health Monitoring Techniques
  • Silicon Carbide Semiconductor Technologies
  • Structural Engineering and Vibration Analysis
  • Fluid Dynamics and Vibration Analysis
  • Fluid Dynamics and Turbulent Flows
  • Computational Fluid Dynamics and Aerodynamics
  • Topic Modeling
  • Natural Language Processing Techniques
  • Microgrid Control and Optimization
  • Wind and Air Flow Studies
  • Aerodynamics and Acoustics in Jet Flows
  • Railway Engineering and Dynamics

Nankai University
2025

China Academy of Launch Vehicle Technology
2020

Central South University
2012

Zhejiang University
2005-2012

A new topology with fault-tolerant ability that improves the reliability of multilevel converters is proposed. This developed through analysis different power device's failure modes. With proposed scheme, function stage can be maintained even part it fails. Its results from redundant nature multiswitching-state and control signal modification. Furthermore, balances voltage levels automatically without any assistance other circuits. The validity scheme confirmed by experiments in a five-level...

10.1109/tpel.2004.842983 article EN IEEE Transactions on Power Electronics 2005-03-01

A non-isolated zero-voltage transition interleaved high step-up converter with built-in transformer is proposed for and large current applications. The consists of an structure, a two sets active clamp circuits. can provide voltage conversion ratio increase the system efficiency without extreme duty cycle compared conventional boost converter. leakage inductance removes output diode reverse-recovery problem, which results in low losses. scheme suppress surge power MOSFETs recycle energy,...

10.1049/iet-pel.2010.0133 article EN IET Power Electronics 2011-05-06

The ability of processing long contexts is crucial for large language models (LLMs), but training LLMs with a long-context window requires substantial computational resources. Many sought to mitigate this through the sparse attention mechanism. However, faces noticeable gap compared full in capturing long-distance information, leading limited capabilities. To effectively address issue, article proposes novel transformer architecture called 2-D (2D-former), aimed at extending context windows...

10.1109/tnnls.2025.3548047 article EN IEEE Transactions on Neural Networks and Learning Systems 2025-01-01

In this study, an advanced interleaved active clamp dual-boost converter is proposed for distributed generation systems with large input current, high-output-voltage and high-step-up conversion. There are only two coupled inductors in the both of them can be integrated into one magnetic core to improve power density. The multi-functional serve as current ripple cancelation transformer primary–secondary isolation. Furthermore, improved voltage-doubler configuration derived on secondary side...

10.1049/iet-pel.2012.0031 article EN IET Power Electronics 2012-08-01

Numerical simulation of flow-field has become an indispensable tool for aerodynamic design. Usually, wall surface integration is a used to calculate values pressure drag and skin friction drag, but the mechanism production still confusing. In present work, in order decompose total into viscous wave induced spurious far-field decomposition (FDD) method developed. This depends on axial velocity defect area sensor functions. The work proposes three hybrid formulas tackle negative square root...

10.1177/0954410020973904 article EN Proceedings of the Institution of Mechanical Engineers Part G Journal of Aerospace Engineering 2020-12-08
Coming Soon ...