Haifeng FANG

Dr. Haifeng FANG


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
25 April 2024 in Ocean Engineering
Verified
Journal Article
Closed access
JFJihua Fan
Jihua Fan
HBHonglin Bai
Honglin Bai
MWMingqiang Wang
Mingqiang Wang
HFHaifeng FANG
Haifeng FANG

18 March 2022 in Computational Intelligence and Neuroscience
Verified
Journal Article
Open access
PDF file
HFHaifeng FANG
Haifeng FANG
JCJin Cao
Jin Cao
ZLZhi-Yuan Li
Zhi-Yuan Li
One of a very significant computer vision task in many real-world applications is traffic sign recognition. With the development of deep neural networks, state-of-art performance traffic sign recognition has been provided in recent five years. Getting very high accuracy in object classification is not a dream any more. However, one of the key challenges is becoming making the deep neural network suitable for an embedded system. As a result, a small neural network with as less parameters as possible and high accuracy needs to be explored. In this paper, the MicronNet which is a small but powerful convolutional neural network is improved by batch normalization and factorization, and the proposed MicronNet-BN-Factorization (MicronNet-BF) takes advantages about reducing parameters and improving accuracy. The effect of image brightness is reduced for feature recognition by the elimination of mean and variance of each input layer in MicronNet via BN. A lower number of parameters are realized with the replacement of convolutional layers in MicronNet, which is the inspiration of factorization. In addition, data augmentation is also been changed to get higher accuracy. Most important, the experiment shows that the accuracy of MicronNet-BF is 99.383% on German traffic sign recognition benchmark (GTSRB) which is much higher than the original MicronNet (98.9%), and the most influence factor is batch normalization after the confirmation of orthogonal experimental. Furthermore, the handsome training efficiency and generality of MicronNet-BF indicate the wide application in embedded scenarios.

20 November 2025 in Applied Thermal Engineering
Verified
Journal Article
Closed access
HFHaifeng FANG
Haifeng FANG
YCYujie Chen
Yujie Chen
LCLihua Cai
Lihua Cai
JYJiang Yao
Jiang Yao
ZZZiwei Zhang
Ziwei Zhang

1 January 2024 in Thermal Science
Verified
Journal Article
Open access
PDF file
HFHaifeng FANG
Haifeng FANG
JXJuncheng Xu
Juncheng Xu
YLYiqiang Lu
Yiqiang Lu
MWMingqiang Wang
Mingqiang Wang
XLXin Li
Xin Li
In order to investigate the impact of cooling plate channel structural parameters on the cooling performance of battery modules, a heat generation model for LiFePO4 batteries was established. Based on the model, the 1C discharge process of LiFePO4 batteries at room temperature (25?) was simulated, and relevant heat release data were obtained. On this basis, three different cooling plate structures (Model A, Model B, Model C) were designed, and the cooling performance of the cooling liquid (50% water and 50% ethylene glycol) for the battery module was analyzed by simulation at different mass-flow rates (0.15 kg/s, 0.18 kg/s, 0.21 kg/s), along with the pressure, temperature difference and flow rate of the cooling channel. The results showed that the uniformity of the flow rate in the channel can reduce the temperature difference. Under the same mass-flow rate, the temperature difference of the battery module on the same surface between Model A and Model C was 1.1?, but too many channels would increase the pressure drop. The pressure drop of Model C was more than 10 times that of Model B. Therefore, it is necessary to design the channel structure reasonably while ensuring the heat dissipation effect. Finally, based on the simulation results, beneficial suggestions for the cooling and cooling plate design and manufacture of energy storage container battery modules are proposed.

1 January 2021 in Thermal Science
Verified
Journal Article
Open access
PDF file
HFHaifeng FANG
Haifeng FANG
JCJin Cao
Jin Cao
KXKanwen Xu
Kanwen Xu
LCLihua Cai
Lihua Cai
HWHaoda Wu
Haoda Wu
QWQubiao Wu
Qubiao Wu
For the sake of comprehend the influence of locally enhanced heat dissipation of the polar on Li-ion power batteries, coupling the thermal effect of anode and cathode, the heat generation model of Li-ion battery with different discharge magnification is obtained. Based on this model, the discharge process simulation analysis of the single battery is carried out and compared with the experimental results. The experiment results show that battery polarity heat effect has great influence on the temperature field distribution of the battery. Then, a locally enhanced heat dissipation structure is set up near the polar region and a comparative experiment is carried out by changing the discharge rate. Although the heat pipe can only slightly improve the discharge capacity of lithium ion battery by enhancing the heat dissipation capability of polarity, the heat generated under low discharge rate can be reduced by using the heat pipe with low emission rate, the heat generated by the polar region can be effectively and timely exported to reduce the temperature of the polar region and it can greatly reduce the overall temperature of the battery. Then, according to the thermal characteristics and the results of locally enhanced heat dissipation analysis, a new battery module is designed and simulated. And the prototype is completed and tested. The results show that this method can effectively reduce the temperature rise and temperature difference of the battery module by using locally enhanced heat dissipation structure. Finally, combined with the simulation and the experimental results, some useful suggestions are put forward for the design and manufacture of battery modules and battery boxes.

7 May 2024 in International Journal of Clothing Science and Technology
Verified
Journal Article
Closed access
HFHaifeng FANG
Haifeng FANG
JZJun Zhang
Jun Zhang
HSHanlin Sun
Hanlin Sun
LCLihua Cai
Lihua Cai
Purpose: As a new type of spinning machine, the jet spinning machine absorbs the carding system of the rotating cup spinning series and the nozzle part of the jet spinning. This paper aims to intends to introduce the double carding structure currently studied by the rotating cup spinning into the jet spinning machine, and analyze the influence of the nozzle characteristic number on the flow field in the double carding structure to verify the advantages of the double carding structure.Design/methodology/approach: The simulation is used to evaluate the performance of single/double split jet spinning and nozzle feature number, verify the technical advantages of double split jet spinning and evaluate the influence of nozzle feature number on flow field. The influence of the nozzle characteristic number on the flow pattern in the four models is compared. The advantages and disadvantages of a conventional single comb and a double comb with a bypass channel on the longer side of the transport channel as an additional air supply channel are also evaluated.Findings: At present, the double comb technology of rotary cup spinning is being studied at home and abroad to improve the spinning quality and improve the difficult problem of mixed yarn with large difference in processing fiber properties. At present, the jet spinning machine combines the advantages of rotary cup spinning and jet spinning, absorbing the comb system of rotary cup spinning series and the nozzle part of jet spinning. Therefore, it is found that the introduction of the double-split structure into the wool jet spinning has research value to improve the spinning quality.Originality/value: The purpose of this paper is to refer to the previous research on the double comb structure in rotary spinning, and to apply the double comb structure in the new jet spinning machine to improve the spinning quality. The simulation is used to evaluate the performance of single/double split jet spinning and nozzle feature number, verify the technical advantages of double split jet spinning and evaluate the influence of nozzle feature number on flow field.