Junkun Zhang

Junkun Zhang

Auto-generated profile, based on publicly available information

Junkun Zhang

College of Electrical and Information Engineering, Hunan University, Changsha, China

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
14 July 2021 in IEEE Transactions on Industry Applications
Verified
Journal Article
Closed access
LWLei Wang
Lei Wang
UMUdaya Kumara Madawala
Udaya Kumara Madawala
JZJunkun Zhang
Junkun Zhang
MWMan-Chung Wong
Man-Chung Wong
There is a growing demand in the marketplace for wirelessly powered consumer applications. Some of these low-power applications require wireless power transfer with bidirectional power flow. For such consumer applications, this article proposes a new bidirectional wireless power transfer (BD-WPT) topology that facilitates wireless power transfers in both directions using fewer components. This article describes the operating principles of the topology in detail and presents a mathematical model to predict and investigate its performance. Besides, a control method with synchronization is also proposed for this topology to regulate power transfer in both directions. Simulations and theoretical results are presented in comparison to the measured results of a 60-W prototype BD-WPT to demonstrate the viability of the proposed converter topology. Results are convincing and indicate that the proposed converter is suitable for low-power applications being low cost, reliable, and efficient.

1 March 2022 in IEEE Transactions on Power Electronics
Verified
Journal Article
Closed access
JZJunkun Zhang
Junkun Zhang
BGBing Gao
Bing Gao
ZHZhixing He
Zhixing He
LWLei Wang
Lei Wang
RHRenjie Hou
Renjie Hou
RHRong Han
Rong Han
High-efficiency high voltage dc power supply is the core equipment for magnetron in the microwave industry. This article presents an input-parallel output-serial modular high isolated resonant converter for the magnetron. However, the high stray capacitance of the high isolation transformer brings some challenges, such as high turn- off current, longer dead time, increasing gain, and transformer optimization. A simplified model including the stray capacitances is built to deal with these problems, and the minimum turn- off current is derived after the detailed time-domain analysis. The appropriate deadtime and the maximum magnetic inductance needed for the proposed converter are obtained with the minimum turn- off current. Then, a 15 kW, 60 kV isolation multiple-windings output transformer is designed and optimized for the converter. The insulation design, core shape, losses, and parasitic parameters are calculated thoroughly. The Pareto optimization process optimizes the switching frequency and the transformer's turns to obtain higher efficiency and a more stable gain. The finite element method and time-domain simulation verified the optimized design results. Finally, a 0.8/2.5 kV, 15 kW all-SiC dc–dc converter module is developed to validate the proposed design. The results indicate that the module efficiency is as high as 98.6%.

1 March 2023 in IEEE Transactions on Industry Applications
Verified
Journal Article
Closed access
JZJunkun Zhang
Junkun Zhang
KMKai Ma
Kai Ma
ELErtao Lei
Ertao Lei
RHRong Han
Rong Han
YLYang Liu
Yang Liu
YMYanjun Ma
Yanjun Ma
Modular input-parallel output-serial (IPOS) DC-DC converter is becoming a crucial solution for high-voltage applications due to its flexible combination form without switching device pressure limitation. It requires high efficiency, fast response, and wide-range output for broader applications. However, the existing solutions cannot simultaneously meet all these requirements. This paper proposes a new hybrid structure with interleaved LLC converters and a DAB converter. The interleaved LLC converters provide the most output voltage with fixed switching frequency and duty cycle, while the DAB converter regulates the output voltage. The wide range output is achieved by switching the LLC converter module and continuously adjusting the output voltage of the DAB converter. Furthermore, a switching control method based on the sliding-mode variable structure is proposed to achieve a smooth and fast output voltage response during the switching process. Finally, a 0.3/0.9-1.8kV, 2kW experimental platform is established to verify the proposed converter's effectiveness and advantages.

2 November 2021 in IEEE Journal of Emerging and Selected Topics in Power Electronics
Verified
Journal Article
Closed access
RHRong Han
Rong Han
QXQianming Xu
Qianming Xu
HDHongqi Ding
Hongqi Ding
BGBing Gao
Bing Gao
LWLei Wang
Lei Wang
ALAn Luo
An Luo
The extremely low-frequency power amplifier (ELFPA) used in the communication system is required to have high output accuracy, small size, and high reliability under complex mission profiles simultaneously. However, the system parameters that determine these performances are coupled with each other, and it is difficult to directly select the optimal solution that meets all output performances at the same time. Thus, a new high reliable parameters design method is proposed for ELFPA system, which can simultaneously consider the relationship among multiple mutually coupled, nonlinear control variables and output performances. Moreover, the multiobjective particle swarm optimization (MOPSO) algorithm is used to search for the Pareto-optimal solutions and the best design. The use of this optimization algorithm aims to avoid the empirical trial-and-error method for adjusting the parameter values and quickly find the system parameter value that best trades off the total harmonic distortion (THD), volume, and lifetime of the system. To illustrate the superiority of the proposed approach, the proposed parameter design process of a 5-kW power amplifier is verified by both simulation and experimental results.

10 May 2021 in IEEE Journal of Emerging and Selected Topics in Power Electronics
Verified
Journal Article
Closed access
JZJunkun Zhang
Junkun Zhang
ZHZhixing He
Zhixing He
YLYang Liu
Yang Liu
ALAn Luo
An Luo
LWLei Wang
Lei Wang
YPYing Pang
Ying Pang
This paper presented a high-efficiency push-pull resonant converter solution for auxiliary power supply in 70 kV isolated applications. This power supply consists of a two-stage structure with a resonant push-pull inverter and a buck regulator. A series resonance is introduced in the inverter with its operation and parameter design to achieve high-efficiency voltage isolation. A buck converter with feedforward control is presented with the small-signal model and controller design to regulate the output voltage. Besides, the transformer is another challenge as a crucial part of the voltage isolation. Specifically, the optimization design of toroidal core and sector windings is presented, including the analysis of isolation, loss analysis, and leakage inductance. The paper presents comprehensive converter and transformer models, describing the proposed power supply’s design process. Furthermore, a case study is demonstrated under 70 kV isolation with 36 V-84 V input and 48 V output. A prototype of a 200 W, 1.4 L isolated power supply has been built to demonstrate the design model’s validity and process. Results convincingly indicate that the proposed converter has the highest efficiency of 93.1% and can maintain the 70 kV isolation voltage with good regulation performance.

1 June 2021
Verified
Conference Paper
Closed access
JTJunfeng Tian
Junfeng Tian
JZJunkun Zhang
Junkun Zhang
FLFanming Liu
Fanming Liu
The automatic detection of software vulnerabilities is a very active research area in software security. Existing machine learning-based vulnerability detectors have been able to attain improved detection capabilities, but they face the problem of a coarse detection granularity or low positioning accuracy. In this paper, we introduce the bounding box regression method to vulnerability detection by referring to the methods and ideas in the target detection field. At the same time, combined with the vulnerability detection based on deep learning, we propose and design a fine-grained vulnerability detection system called BBregLocator. The system uses bounding boxes to mark the locations of vulnerabilities, can simultaneously detect and locate vulnerabilities, and improves the accuracy of vulnerability location through bounding box regression. The experimental results show that BBregLocator can effectively improve the vulnerability location accuracy while maintaining low false-positive and false-negative rates.